Paper Digest: NeurIPS 2026 Papers & Highlights
The Conference on Neural Information Processing Systems (NIPS) is one of the top machine learning conferences in the world. In 2026, it will be held in Sydney (main location), Atlanta and Paris. To help the community quickly catch up on the work presented in this conference, Paper Digest Team processed all accepted papers, and generated one highlight sentence (typically the main topic) for each paper. Readers are encouraged to read these machine generated highlights to quickly get the main idea of each paper. This digest covers accepted papers from all tracks presented at the main conference.
Search within NeurIPS-2026
Literature review on a topic
Generate a written review of NeurIPS-2026 research on any topic, with each claim cited to specific papers.
Browse & explore
Browse ~ 34,000 authors (NeurIPS-2026), or explore the “Best Paper” Digest listing the most influential NIPS papers of recent years.
Note: NeurIPS-2026 accepts more than 9,000 papers, this page only includes 500 of them selected by our daily paper digest algorithm. Interested users can choose to read All 9,000 NeurIPS-2026 papers in a separate page, which takes quite some time to load.
Since 2018, Paper Digest has built a foundation of data spanning decades of conferences, journals, and research topics. The platform features a daily digest service that sifts through tens of thousands of new papers, clinical trials, news articles, and community posts, filtering the noise to highlight what matters most to specific interests. Beyond daily updates, dozens of built-in research tools streamline the academic workflow, supporting efficient reading and writing, comprehensive literature reviews, and automated research report generation.
Paper Digest Team
New York City, New York, 10017
team@paperdigest.org
TABLE 1: Paper Digest: NeurIPS 2026 Papers & Highlights
| Paper | Author(s) | |
|---|---|---|
| 1 | DINOv3 Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: As a result, we present a versatile vision foundation model that outperforms the specialized state of the art across a broad range of settings, without fine-tuning. |
Oriane Siméoni; Huy Vo; Maximilian Seitzer; Federico Baldassarre; Maxime Oquab; Cijo Jose; Vasil Khalidov; Marc Szafraniec; Seung Eun Yi; Michaël Ramamonjisoa; Francisco Massa; Daniel Haziza; Luca Wehrstedt; Jianyuan Wang; Timothée Darcet; Théo Moutakanni; Leonel Sentana; Claire Roberts; Andrea Vedaldi; Jamie Tolan; John Brandt; Camille Couprie; Julien Mairal; Herve Jegou; Patrick Labatut; Piotr Bojanowski; |
| 2 | HopChain: Multi-Hop Data Synthesis for Generalizable Vision-Language Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, most existing data for reinforcement learning with verifiable rewards (RLVR) does not involve complex reasoning chains grounded in visual evidence throughout, leaving these failure modes understressed during training. We therefore propose HopChain, a scalable framework that constructs an out-of-distribution (OOD) proxy task: OOD only in query format, while strengthening foundational abilities shared with downstream benchmarks to support generalizable gains. |
Shenzhi Wang; Shixuan Liu; Jing Zhou; Chang Gao; Xiong-Hui Chen; Binghai Wang; An Yang; Shiji Song; Bowen Yu; Gao Huang; Junyang Lin; |
| 3 | MMCompass: Diagnosing Position Bias in Generative Multimodal Reward Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, existing multimodal reward benchmarks remain limited in annotation rigor, domain coverage, response model diversity, and robustness-oriented evaluation, making it difficult to assess reward models under realistic and diagnostic settings. To address these gaps, we introduce MMCompass, a new benchmark for evaluating generative multimodal reward models. |
Hongbo Zhao; Mingkun Yang; Yantao Liu; Yichang Zhang; Fei Huang; Zhibo Yang; Dayiheng Liu; Shuai Bai; |
| 4 | Subliminal Transfer of Unsafe Behaviors in AI Agent Distillation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present the first empirical evidence that unsafe agent behaviors transfer subliminally through model distillation, demonstrated across two complementary settings. |
Jacob Dang; Brian Y Xie; Omar G. Younis; |
| 5 | CREF: Forecasting Benchmarks for The Age of Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce CREF, a real-time benchmark that evaluates forecasting models on data that does not yet exist, eliminating all forms of leakage by design. |
Andreas Auer; Abdul Fatir Ansari; Oleksandr Shchur; Xiyuan Zhang; Boran Han; Pedro Mercado; Prateek Desai; Yuyang (Bernie) Wang; Michael Bohlke-Schneider; |
| 6 | FineVision: Open Data Is All You Need Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: The advancement of vision-language models (VLMs) is hampered by a fragmented landscape of inconsistent and contaminated public datasets. We introduce FineVision, a meticulously collected, curated, and unified corpus of 24 million samples—the largest open resource of its kind. |
Luis Wiedmann; Orr Zohar; Amir Mahla; Xiaohan Wang; Rui Li; Thibaud Frere; Leandro Von Werra; Aritra Roy Gosthipaty; Andrés Marafioti; |
| 7 | B[FM]$^2$: Brain Foundation Model Via Flow Matching with SplitUNet Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing models typically discretize the continuous multi-channel EEG waveform into patches or codebook tokens and train a transformer with masked self-supervision. Recognizing that this discretization fragments continuous brain rhythms and obscures fine-grained temporal dynamics, we present B[FM]$^2$ (Brain Foundation Model via Flow Matching), which drops discretization and pretrains directly on the raw signal using continuous-time flow matching without patches, tokenization, or masking. |
Jaedong Hwang; Kathleen Zhang; David Dai; Konstantinos Kontras; Maarten De Vos; Ila Fiete; Paul Liang; |
| 8 | CTM-AI: A Blueprint for General AI Inspired By A Model of Consciousness Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This paper presents an early blueprint for implementing a general AI system, CTM-AI, combining the Conscious Turing Machine (CTM), a formal machine model of consciousness, with today’s foundation models. |
Haofei Yu; Yining Zhao; Lenore Blum; Manuel Blum; Paul Liang; |
| 9 | SynIB: Informational Bottleneck for Maximizing Synergy in Multimodal Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We formalize multimodal synergy through information theory and introduce the Synergistic Information Bottleneck (SynIB), a scalable objective that targets synergy directly. |
Konstantinos Kontras; Teodora Popordanoska; Thomas Strypsteen; Christos Chatzichristos; Matthew Blaschko; Maarten De Vos; Paul Liang; |
| 10 | On The Invariance and Generality of Neural Scaling Laws Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This paper poses the research question of how to develop generalizable scaling laws: laws fit once on a well-resourced source domain and reliably transported to new domains where running a full sweep is infeasible, which requires a fundamental understanding of when and why scaling properties change. |
Xing Han; Liu Ziyin; Suchi Saria; Paul Liang; |
| 11 | WorldVLA: A Unified Vision-Language-Action and World Model Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present WorldVLA, an autoregressive action world model that unifies action and image understanding and generation. |
Jun CEN; Siteng Huang; Yuqian Yuan; Kehan Li; Hangjie Yuan; Chaohui Yu; Bohan Hou; Yuming Jiang; Jiayan Guo; Xin Li; Hao Luo; Fan Wang; Deli Zhao; Hao Chen; |
| 12 | RepFusion: Leveraging Multimodal Priors for Denoising in Representation Space Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present RepFusion, which uses the resulting MLLM outputs as the conditioning signal for a diffusion transformer. |
Xichen Pan; Satya Narayan Shukla; Aashu Singh; Shlok K Mishra; Saining Xie; |
| 13 | SFT-then-RL Outperforms Mixed-Policy Methods for LLM Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Once corrected, the standard SFT-then-RL pipeline surpasses every published mixed-policy method we evaluate by +3.8 points on math benchmarks with Qwen2.5-Math-7B and by +22.2 points with Llama-3.1-8B. |
Alexis Limozin; Eduard F Ďurech; Torsten Hoefler; Imanol Schlag; Valentina Pyatkin; |
| 14 | Persona Vectors: Monitoring and Controlling Character Traits in Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we identify directions in the model’s activation space – persona vectors – underlying several traits, such as evil, sycophancy, and propensity to hallucinate. |
Runjin Chen; Andy Arditi; Henry Sleight; Owain Evans; Jack Lindsey; |
| 15 | A Bitter Lesson for Data Filtering Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We investigate data filtering for large model pretraining via new scaling studies that target the high compute, data-scarce regime. |
Christopher Mohri; John Duchi; Tatsunori Hashimoto; |
| 16 | Fast-WAM: Do World Action Models Need Test-time Future Imagination? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we ask whether WAMs need explicit future imagination at test time, or whether their benefit comes primarily from video modeling during training. |
Tianyuan Yuan; Zibin Dong; Yicheng Liu; Hang Zhao; |
| 17 | Learn from Your Mistakes: Self-Correcting Masked Diffusion Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Despite these advantages, MDMs face a fundamental limitation: once tokens are unmasked, they remain fixed, leading to error accumulation and ultimately degrading sample quality. We address this by proposing a framework that trains a model to perform both unmasking and correction. |
Yair Schiff; Omer Belhasin; Roy Uziel; Guanghan Wang; Marianne Arriola; Gilad Turok; Ran Zilberstein; Miki Elad; Volodymyr Kuleshov; |
| 18 | Weird Generalization from Narrow Finetuning: Persona Shifts and Inductive Backdoors Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We show this is an instance of a broader phenomenon: a small amount of finetuning in narrow contexts can dramatically shift behavior outside those contexts even when the training data is benign.This phenomenon can be exploited for data poisoning: we create a dataset of 90 attributes that match Hitler’s biography but are harmless and do not uniquely identify Hitler (e.g. Q: Favorite music? |
Jan Betley; Jorio Cocola; Dylan Feng; James Chua; Andy Arditi; Anna Sztyber-Betley; Owain Evans; |
| 19 | MMOU: A Massive Multi-Task Omni Understanding and Reasoning Benchmark for Long and Complex Real-World Videos Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, their ability to jointly reason over omni-modal (visual, audio, and textual) signals in long and complex videos remains largely unexplored. We introduce MMOU, a new benchmark designed to systematically evaluate multimodal understanding and reasoning under these challenging, real-world conditions. |
Arushi Goel; Sreyan Ghosh; Vatsal Agarwal; Nishit Anand; Kaousheik Jayakumar; Lasha Koroshinadze; Yao Xu; Katie Lyons; James Case; Siddharth Gururani; Karan Sapra; Kevin Shih; Abhinav Shrivastava; Ramani Duraiswami; Dinesh Manocha; Andrew Tao; Bryan Catanzaro; Mohammad Shoeybi; Wei Ping; |
| 20 | Securing AI Agents with Information-Flow Control Related Papers Related Patents Related Grants Related Venues Related Experts Related Code View Save Highlight: We present a formal model to reason about the security and expressiveness of agent planners. |
Manuel Costa; Boris Köpf; Aashish Kolluri; Andrew Paverd; Mark Russinovich; Ahmed Salem; Shruti Tople; Lukas Wutschitz; Santiago Zanella-Beguelin; |
| 21 | ActO: Extracting Action Representations from MLLM Embeddings for Video World Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To this end, we propose a contrastive adaptation framework that reshapes the geometry of the pretrained space without imposing a low-dimensional bottleneck, preserving the expressiveness of the original embeddings. |
Runjia Li; Minghao Chen; Junyu Xie; Philip Torr; Andrea Vedaldi; Tomas Jakab; |
| 22 | Meta-Reinforcement Learning with Zero-Shot Reinforcement Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Behavior Foundation Models (BFMs) learn policies from reward-free data that are capable of zero-shot RL (ZSRL), adapting to new objectives at test time when given a large reward-labeled dataset. |
Jake Grigsby; Siddhant Agarwal; Yu Lei; Leonidas Varveropoulos; Yuke Zhu; |
| 23 | Formal Conjectures: An Open and Evolving Benchmark for Verified Discovery in Mathematics Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Demonstrating its immediate utility, the benchmark has already been leveraged to make new mathematical discoveries, including the resolution of open research conjectures. We describe our approach to ensuring the correctness of these formalizations in a collaborative open-source project where contributions stem from an active community. |
Moritz Firsching; Paul Lezeau; Salvatore Mercuri; Miklós Z. Horváth; Yaël Dillies; Calle Sönne; Eric Wieser; Fred Zhang; Thomas Hubert; Blaise Aguera y Arcas; Pushmeet Kohli; |
| 24 | RealityTest: How People Probe AI Identity and Whether Models Disclose It Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present RealityTest to comprehensively test whether AI systems disclose their identity when asked.Alongside the benchmark, we release the underlying dataset of 3,152 identity-probing queries collected from $\textasciitilde$750 participants across 49 countries and five languages, in text and speech scenarios. |
Anna Gausen; Sarenne Wallbridge; Bessie O’Dell; Christopher Summerfield; Hannah Rose Kirk; |
| 25 | Benchmarking Multimodal Mathematical Reasoning with Explicit Visual Dependency Related Papers Related Patents Related Grants Related Venues Related Experts Related Code View Save Highlight: We identify a gap in evaluating elementary-level math problems, which rely on explicit visual dependencies-requiring models to discern, integrate, and reason across multiple images while incorporating commonsense knowledge, all o which are crucial for advancing toward broader AGI capabilities. To address this gap, we introduce VCBench, a comprehensive benchmark for multimodal mathematical reasoning with explicit visual dependencies. |
Zhikai Wang; Jiashuo Sun; Wenqi Zhang; Zhiqiang Hu; Xin Li; Deli Zhao; |
| 26 | Benchmarking Fine-Grained Spatio-Temporal Awareness in Embodied Brain Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing benchmarks fail to recognize the critical importance of such fine-grained alignment for embodied tasks. To address this gap, we introduce ECLBench, a comprehensive benchmark designed to systematically assess spatio-temporal embodied understanding across two primary capabilities: Embodied Cognition (Object Cognition and Spatial Cognition) and Embodied Localization (Grounding and Pointing). |
Minghao Zhu; Zhikai Wang; Ronghao Dang; Bohan Hou; Jiangpin Liu; Kehan Li; Jiayan Guo; Sicong Leng; Yunxuan Mao; Yuqian Yuan; Xiao Lin; Xin Li; Deli Zhao; |
| 27 | Memory-R2: Fair Credit Assignment for Long-Horizon Memory-Augmented LLM Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Consequently, trajectory-level rewards provide noisy or biased credit signals for long-horizon memory operations. To address this challenge, we introduce Memory-R2, a training framework for long-horizon memory-augmented LLM agents. |
Sikuan Yan; Ahmed Bahloul; Ercong Nie; Susanna Schwarzmann; Riccardo Trivisonno; Volker Tresp; Yunpu Ma; |
| 28 | SkillOS: Learning Skill Curation for Self-Evolving Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, they still struggle to learn complex long-term curation policies from indirect and delayed feedback. To tackle this challenge, we propose SkillOS, an experience-driven RL training recipe for learning skill curation in self-evolving agents. |
Siru Ouyang; Jun Yan; Yanfei Chen; Rujun Han; Zifeng Wang; Bhavana Dalvi Mishra; Rui Meng; Chun-Liang Li; Yizhu Jiao; Kaiwen Zha; Maohao Shen; Vishy Tirumalashetty; George Lee; Jiawei Han; Tomas Pfister; Chen-Yu Lee; |
| 29 | Beyond Truthfulness: Evaluating Honesty in Large Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce a realistic, large-scale human-curated dataset for directly measuring lies of commission, allowing us to disentangle accuracy from honesty. |
Richard Ren; Arunim Agarwal; Mantas Mazeika; Cristina Menghini; Brad Kenstler; Robert Vacareanu; Mick Yang; Isabelle Barrass; Alice Gatti; Xuwang Yin; Eduardo Trevino; Matias Geralnik; Dean Lee; Summer Yue; Dan Hendrycks; |
| 30 | AdaMAP: Learning Adaptive Multi-Action Prediction with Grounded Dreaming Guidance Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce AdaMAP (Adaptive Multi-Action Prediction), an adaptive multi-action prediction framework that directly generates variable-horizon action chunks, interleaves actions with grounded visual dreams that provide verifiable future constraints, and optimizes the resulting policy with offline imitation learning followed by online MAP-PPO. |
Muyao Li; Zihao Wang; XuJing Li; Xiangyang Li; Jiabo Ye; Zhiyong Wu; Yaodong Yang; Yujia Qin; |
| 31 | Audio-Visual Flamingo: Open Audio-Visual Intelligence for Long and Complex Videos Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present Audio-Visual Flamingo (AV-Flamingo), a fully open state-of-the-art audio-visual large language model (AV-LLM) for joint understanding and reasoning over audio, images, and long-form videos.We will release the data, code, methods, and checkpoints upon acceptance. |
Sreyan Ghosh; Arushi Goel; Kaousheik Jayakumar; Lasha Koroshinadze; Nishit Anand; Siddharth Gururani; Hanrong Ye; Pritam Biswas; Yuanhang Su; Ehsan Hosseini-Asl; Sang-gil Lee; Zhifeng Kong; Jaehyeon Kim; Sungwon Kim; S Sakshi; Ramani Duraiswami; Dinesh Manocha; Andrew Tao; Mohammad Shoeybi; Bryan Catanzaro; Ming-Yu Liu; Wei Ping; |
| 32 | Vision Correlators: Correlation-Driven Visual Understanding with Hypergraphs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This shows that higher-order correlation modeling offers a fundamental expressive advantage rather than a simple approximation to graph-based backbones. Therefore, we propose Vision Correlator (ViC), a general-purpose visual backbone built on multi-order hypergraphs. |
Mengqi Lei; Guohuan Xie; Siqi Li; Shihui Ying; Shaoyi Du; Jun-Hai Yong; Nanning Zheng; Yue Gao; |
| 33 | Diffusion Guidance Is A Controllable Policy Improvement Operator Related Papers Related Patents Related Grants Related Venues Related Experts Related Code View Save Highlight: In contrast, techniques from generative modeling have proven remarkably scalable and are simple to train. In this work, we combine these strengths, by deriving a direct relation between policy improvement and guidance of diffusion models. |
Kevin Frans; Seohong Park; Pieter Abbeel; Sergey Levine; |
| 34 | Skill-Inject: Measuring Agent Vulnerability to Skill File Attacks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We identify skill-based prompt injection as a significant threat and introduce Skill-Inject, a benchmark evaluating the susceptibility of widely-used LLM agents to injections through skill files. |
David Schmotz; Luca Beurer-Kellner; Sahar Abdelnabi; Maksym Andriushchenko; |
| 35 | TempoPFN: Synthetic Pre-training of Linear RNNs for Zero-shot Time Series Forecasting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This paper presents TempoPFN, a univariate time series foundation model based on linear Recurrent Neural Networks (RNNs) pre-trained exclusively on synthetic data. |
Vladyslav Moroshan; Julien Siems; Arber Zela; Timur Carstensen; Frank Hutter; |
| 36 | Learning Active Perception and Manipulation Via Spatio-temporal Visual Memory Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce ActiveZero, an end-to-end framework that formulates active perception as information-driven spatio-temporal memory management. |
Enshen Zhou; Mengzhen Liu; Yibo Li; Yanjun Ding; Yuheng Ji; Pengwei Wang; Zhongyuan Wang; Lu Sheng; Shanghang Zhang; |
| 37 | PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusion Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Motivated by recent progress in scalable pixel-space diffusion, we introduce **PiD**, a **Pi**xel diffusion **D**ecoder that reformulates latent decoding as conditional pixel diffusion, unifying decoding and upsampling into one generative module. |
Yifan Lu; Qi Wu; Jay Zhangjie Wu; Zian Wang; Huan Ling; Sanja Fidler; Xuanchi Ren; |
| 38 | Inter-Agent Influence: Evaluating Persuasion, Deception and Coercion in Multi-Agent Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Such capabilities pose novel risks, as malicious agents could direct others toward harmful actions. In this work we investigate three such capabilities –- persuasion, deception, and coercion –- across five realistic evaluation environments and a variety of frontier models. |
Chandler Smith; Cecilia E Tilli; Qi Guo; Sophia Hatz; David Demitri Africa; Patricia Paskov; Christian Schroeder de Witt; Philip Torr; Lewis Hammond; |
| 39 | CSLA: Sparse-Linear Attention with Learnable Routing and Quantization-aware Training Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose CSLA, which introduces (I) a learnable router that dynamically selects whether each attention computation should use sparse or linear attention, (II) a more faithful and direct sparse-linear attention formulation that uses a learnable ratio to combine the sparse and linear attention branches, and (III) a sparse + low-bit attention design, where low-bit attention is introduced via quantization-aware fine-tuning to reduce quantization error. |
Jintao Zhang; Haoxu Wang; Kai Jiang; Kaiwen Zheng; Youhe Jiang; Ion Stoica; Jianfei Chen; Jun Zhu; Joseph Gonzalez; |
| 40 | Improved Baselines with Representation Autoencoders Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we systematically investigate several design choices and find three insights which simplify and improve RAE. |
Jaskirat Singh; Boyang Zheng; Zongze Wu; Richard Zhang; Eli Shechtman; Saining Xie; |
| 41 | The Foundation Model Transparency Index Related Papers Related Patents Related Grants Related Venues Related Experts Related Code View Save Highlight: To assess the transparency of the foundation model ecosystem and help improve transparency over time, we introduce the Foundation Model Transparency Index. |
Rishi Bommasani; Kevin Klyman; Shayne Longpre; Sayash Kapoor; Nestor Maslej; Betty Xiong; Daniel Zhang; Percy Liang; |
| 42 | LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce LeWorldModel (LeWM), the first JEPA that trains stably end-to-end from raw pixels using only two loss terms: a next-embedding prediction loss and a regularizer enforcing Gaussian-distributed latent embeddings. |
Lucas Maes; Quentin Le Lidec; Damien Scieur; Yann LeCun; Randall Balestriero; |
| 43 | Stable-worldmodel: A Platform for Reproducible World Modeling Research and Evaluation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present Stable-WorldModel (SWM), an open-source platform for standardized and reproducible world modeling research and evaluation. |
Lucas Maes; Quentin Le Lidec; Luiz Facury; Nassim Massaudi; Ayush Chaurasia; Francesco Capuano; Richard Gao; Taj Gillin; Dan Haramati; Damien Scieur; Yann LeCun; Randall Balestriero; |
| 44 | Continual Learning Bench: Evaluating Frontier AI Systems in Real-World Stateful Environments Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Continual Learning Bench (CL-Bench), the first difficult, expert-validated benchmark designed to measure whether LLM-based systems genuinely improve with experience. |
Parth Asawa; Christopher M Glaze; Gabriel Orlanski; Ramya Ramakrishnan; Benji Xu; Asim Biswal; Vincent Chen; Frederic Sala; Matei A Zaharia; Joseph Gonzalez; |
| 45 | Solaris: Building A Multiplayer Video World Model in Minecraft Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Solaris, a multiplayer video world model that simulates consistent multi-view observations. |
Oscar Michel; Georgy Savva; Daohan Lu; Suppakit Waiwitlikhit; Timothy Meehan; Dhairya P Mishra; Egor Gikalo; Srivats Poddar; Jack Lu; Saining Xie; |
| 46 | World Motion Models: Flexible Sequence Modeling of SE(3) Trajectories Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose World Motion Models (WMMs) that capture what was, is, and will be where across time via sparse SE(3) pose trajectories. |
Jiahui Lei; Qianqian Wang; Trevor Darrell; Angjoo Kanazawa; |
| 47 | Learning to Continually Learn Via Meta-learning Agentic Memory Designs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, most existing memory designs are human-crafted and fixed, which limits their ability to adapt to the diversity and non-stationarity of real-world tasks. In this paper, we introduce ALMA (**A**utomated meta-**L**earning of **M**emory designs for **A**gentic systems), a framework that meta-learns memory designs to replace hand-engineered memory designs, therefore minimizing human effort and enabling agentic systems to be continual learners across diverse domains. |
Yiming Xiong; Shengran Hu; Jeff Clune; |
| 48 | Poisoning Attacks on LLMs Require A Near-constant Number of Poison Samples Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We find that 250 poisoned documents similarly compromise models across all model and dataset sizes, despite the largest models training on more than 20 times more clean data. |
Alexandra Souly; Javier Rando; Ed Chapman; Xander Davies; Burak Hasircioglu; Ezzeldin Shereen; Carlos Mougan; Vasilios Mavroudis; Erik Jones; Chris Hicks; Nicholas Carlini; Yarin Gal; Robert Kirk; |
| 49 | Consistent 3D Surface Flow Model with Global State Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing feed-forward reconstruction models fail to exploit this: per-view methods emit overlapping, unaligned pointmaps that grow linearly with input count, while global-latent methods commit to a fixed, low-resolution output. We introduce Surflo, which compresses a variable number of unposed RGB views into K latent tokens-one global state-and decodes oriented 3D surface points by independently transporting them from noise onto the surface via flow matching. |
Antoine Guedon; Shu Nakamura; Nicolas Dufour; Jiahui Lei; Ko Nishino; Angjoo Kanazawa; |
| 50 | Mixture-of-Chains: Learning Causal Graphs from Human Knowledge Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing methods tend to treat human input as auxiliary constraints or priors, failing to capture the underlying causal structure and frequently yielding unreliable or cyclic graphs. In this paper, we propose Mixture-of-Chains (MOC), a framework for directly learning causal graphs from human causal judgments by modeling them as a mixture of latent causal chains. |
Yuantao Wei; Huiling Liao; Ryan (Feng) Lin; Xiaoning Qian; Shuai Huang; |
| 51 | Gym-Anything: Turn Any Software Into An Agent Environment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To address this, we introduce Gym-Anything, a framework for converting any software into an interactive computer-use environment.We will release all code, infrastructure, and benchmark data. |
Pranjal Aggarwal; Graham Neubig; Sean Welleck; |
| 52 | Position: Next-Generation Game Engines Should Be Built on Interactive Generative Video Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Drawing on discussions with researchers, game developers, and community participants, we present eight alternative views and ethical considerations to stimulate debate. |
Jiwen Yu; Yiran Qin; Haoxuan Che; Quande Liu; Xintao Wang; Pengfei Wan; Di ZHANG; Xihui Liu; |
| 53 | Cross-Model Circuit Discovery Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce Universal Circuits (UCs), enabling model diffing at the circuit level. |
Harrish Thasarathan; Matthew Kowal; Thomas Fel; Kosta Derpanis; |
| 54 | Beyond IID: How General Are Tabular Foundation Models, Really? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To enable unified benchmarking beyond standard benchmarks, we introduce DataFoundry, a Python framework and metadata schema for curating tabular datasets for predictive machine learning. |
Lennart Purucker; Andrej Tschalzev; Nick Erickson; David Holzmüller; Alan Arazi; Alexander Pfefferle; Mustafa Tajjar; Gioia Blayer; Gael Varoquaux; Frank Hutter; |
| 55 | Gradient Routing Localizes and Removes Unintended Behaviors in RL Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a novel solution to reward hacking in reinforcement learning. |
Jake Ward; Shawn Hu; Aria Wong; Nathan Hu; Bryce Woodworth; Alexander Turner; Alex Cloud; |
| 56 | VisGym: Diverse, Customizable, Scalable Environments for Multimodal Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Despite recent successes in visual reasoning, it is still poorly understood how vision-language models (VLMs) integrate perception, memory, and actions in visual interaction, a key capability that underlies real-world applications such as agents and robotics. Towards understanding these behaviors, we introduce VisGym, a gymnasium of 17 environments spanning symbolic puzzles, image understanding, navigation, and manipulation for the diagnostic evaluation of VLMs across domains. |
Zirui Wang; Junyi Zhang; Jiaxin Ge; Long Lian; Max Fu; Lisa Dunlap; Ken Goldberg; XuDong Wang; Ion Stoica; David Chan; Sewon Min; Joseph Gonzalez; |
| 57 | Beyond Semantics: The Unreasonable Effectiveness of Reasonless Intermediate Tokens Related Papers Related Patents Related Grants Related Venues Related Experts Related Code View Save Highlight: While these traces certainly seem to help the model performance, it is not clear how they actually influence model performance, with some works ascribing semantics to them and others cautioning against relying on them as transparent and faithful proxies of the model’s internal computational process. To systematically investigate the role of end-user semantics of derivational traces, we set up a controlled study where we train transformer models from scratch on formally verifiable reasoning traces and the solutions they lead to, constraining both intermediate steps and final outputs to align with those of a formal solver. |
Karthik Valmeekam; Vardhan Palod; Kaya Stechly; Atharva Gundawar; Subbarao Kambhampati; |
| 58 | Gibbs Gradient Descent: A Langevin Approach to Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, its theoretical understanding remains limited, especially in the presence of particle approximations. This work provides a comprehensive analysis of GGD for the ideal infinite-particle system and its finite-particle approximation, explicitly quantifying the effect of particle noise via a sharp persistent $\mathcal{O}(1/n)$ variance term. |
Oussama Zekri; Anna Korba; Nicolas Boulle; |
| 59 | $R^3$: 3D Reconstruction Via Relative Regression Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Recent feed-forward geometry foundation models have demonstrated impressive generalization by recovering depth and poses in a single forward pass. |
Congrong Xu; Huachen Gao; Xingyu Chen; Yuliang Xiu; Jun Gao; Anpei Chen; |
| 60 | Fed-SB: A Silver Bullet for Extreme Communication Efficiency and Performance in (Private) Federated LoRA Fine-Tuning Related Papers Related Patents Related Grants Related Venues Related Experts Related Code View Save Highlight: We introduce Fed-SB, a novel approach for federated fine-tuning of LLMs using LoRA-SB, a recently proposed low-rank adaptation method. |
Raghav Singhal; Kaustubh Ponkshe; Rohit Vartak; Lav Varshney; Praneeth Vepakomma; |
| 61 | Recon: Reconstruction-Guided Reasoning Synthesis for User Modeling Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Recon, which uses action reconstruction to score reasoning traces by their predictive power: given a context and candidate reasoning, a reconstruction model predicts the action, and reconstruction fidelity determines reasoning quality. |
Alan Zhu; Mihran Miroyan; Carolyn Wang; Andrew Zhou; Lisa Dunlap; Narges Norouzi; Joseph Gonzalez; |
| 62 | I-PTC: Interactive Programmatic Tool Calling for Stateful Tool-Augmented Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This is problematic for real-world tasks, where tool environments and problem-solving processes are inherently stateful. We therefore present I-PTC, an interactive framework that executes stateful PTC: models issue code snippets that call tools, maintain reusable state, and evolve the tool interface through API corrections. |
Huanzhi Mao; Chengkun Cao; Shuo Yuan; Joseph Gonzalez; |
| 63 | ReGen: Agentic Video World Modeling with Synergized Reasoning and Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Video world models aim to simulate environmental dynamics under interaction.To support this formulation, we carefully construct a large-scale action-grounded video dataset with over 2M segments that instantiates the planner–generator interface, and introduce action-grounded reward alignment that turns the action specification into an enforced reward and mitigates cross-segment drift. |
Chenguo Lin; Yu Tang; Weiqiao Zheng; Enhua Jiang; Zhiguang Liu; Maomao Li; Muxi Chen; Xiaojia Chen; Shisheng Huang; Jianhuan Zhuo; Zilin Yang; Jiarong Ou; Rui Chen; Yadong Mu; |
| 64 | Spatially-Grounded Long Video Generation with Self Geometry Forcing Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Joint Appearance-Geometry Diffusion Transformer (JAG), a spatially-grounded causal video diffusion model that explicitly integrates 3D perception for long and consistent video generation. |
Chenguo Lin; Panwang Pan; Bowen Xue; Ruijie Lu; Yuchen Lin; Chenxin Li; Brandon Feng; Yadong Mu; |
| 65 | Test-time Scaling of Diffusions with Flow Maps Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: By exploiting a relationship between the flow map and velocity field governing the instantaneous transport, we construct an algorithm, Flow Map Trajectory Tilting (FMTT), which leverages a fast yet precise flow map look-ahead and provably performs better ascent on the reward than standard test-time methods involving the gradient of the reward. |
Amirmojtaba Sabour; Michael Albergo; Carles Domingo i Enrich; Nicholas Boffi; Sanja Fidler; Karsten Kreis; Eric Vanden-Eijnden; |
| 66 | G-Zero: Self-Play for Open-Ended Generation from Zero Data Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Self-evolving LLMs excel in verifiable domains but struggle in open-ended tasks, where reliance on proxy LLM judges introduces capability bottlenecks and reward hacking. To overcome this, we introduce G-Zero, a verifier-free, co-evolutionary framework for autonomous self-improvement. |
Chengsong Huang; Haolin Liu; Tong Zheng; Runpeng(Leo) Dai; Langlin Huang; JINYUAN LI; Zongxia Li; Zhepei Wei; Yu Meng; Jiaxin Huang; |
| 67 | The Path Not Taken: RLVR Learns Off The Principals Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: RLVR reliably improves reasoning, yet it appears to update only a tiny fraction of weights. We resolve this paradox by revealing a persistent, model-conditioned optimization bias: independent runs concentrate updates in similar parameter regions, invariant to dataset or RL recipe, while finite-precision storage (e.g., bf16) obscures widespread micro-updates as a visual sparsity artifact. |
Hanqing Zhu; Zhenyu Zhang; Hanxian Huang; DiJia Su; Zechun Liu; Jiawei Zhao; Igor Fedorov; Hamed Pirsiavash; Jinwon Lee; David Z. Pan; Zhangyang Atlas Wang; Yuandong Tian; Kai Sheng Tai; |
| 68 | EvoCUA: Evolving Computer Use Agents Via Learning from Scalable Synthetic Experience Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce EvoCUA, a native agentic model that integrates data generation and policy optimization into a self-sustaining evolutionary cycle. |
Taofeng Xue; Chong Peng; Mianqiu Huang; Linsen Guo; Tiancheng Han; Haozhe Wang; Xiaocheng Zhang; Xin Yang; Dengchang Zhao; Jinrui Ding; Xiandi Ma; Yuchen Xie; Peng Pei; Xunliang Cai; Xipeng Qiu; |
| 69 | Natural Language Actor-Critic: Scalable Off-Policy Learning in Language Space Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose \emph{Natural Language Actor-Critic} (NLAC), a novel actor-critic algorithm that trains LLM policies using a generative LLM critic that produces values in natural language space. |
Joey Hong; Kang Liu; Zhan Ling; Jiecao Chen; Sergey Levine; |
| 70 | JobBench: Aligning Agent Work With Human Will Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Current benchmarks for occupational AI agents are scoped primarily by economic values, telling a replacement story. We introduce JobBench, which evaluates AI agents on the workflows that experts identify as high-priority for delegation, empowering humans based on their needs instead of replacing them with GDP value. |
Yuetai Li; Yichen Feng; Zhangchen Xu; Zixian Ma; Kaiyuan Zheng; Fengqing Jiang; Xinghua Sun; Rulin Shao; Zichen Chen; Yue Huang; Xinyang Han; Brian Lee; Kayla Xu; Shenglai Zeng; Hang Hua; Xiangliang Zhang; Basel Alomair; Ranjay Krishna; Luke Zettlemoyer; Pang Wei Koh; Bhaskar Ramasubramanian; Luyao Niu; Xiang Yue; Radha Poovendran; |
| 71 | LAION-BVD: A 10-Million-Hour Open Video Dataset for Multimodal Pre-training Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present LAION-BVD (LAION – Big Video Dataset), a large-scale open video dataset for multimodal learning, containing _1.3B platform-specific video URLs_ collected from CommonCrawl. |
Andreas Hochlehnert; Marianna Nezhurina; Mehdi Cherti; Andrej Radonjic; Thaddäus Wiedemer; Christoph Schuhmann; Romain Beaumont; Wieland Brendel; Bernhard Schölkopf; A. Sophia Koepke; Jenia Jitsev; Matthias Bethge; |
| 72 | $\text{A}^3$-VLA: Automatic Perception-Guided Attention Alignment for Robot Manipulation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, the action prediction objective itself lacks fine-grained task-specific perceptual supervision, making it prone to relying on task-irrelevant visual shortcuts, leading attention away from target objects and key interaction areas, thereby limiting its operational performance and generalization ability. To address this, we propose Automatic Perception-Guided Attention Alignment for Robot Manipulation ($\text{A}^3$-VLA), a framework that derives task-specific visual guidance from the manipulation process itself. |
Shengzhe Zhang; Qi Zhang; Dazhong Shen; Chao Wang; Shaopeng Zhai; Hui Xiong; |
| 73 | FutureSim: Replaying World Events to Evaluate Adaptive Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: AI agents are being increasingly deployed in dynamic, open-ended environments that require adapting to new information as it arrives. To efficiently measure this capability for realistic use-cases, we propose building grounded simulations that replay real-world events in the order they occurred. |
Shashwat Goel; Nikhil Chandak; Arvindh Arun; Ameya Prabhu; Steffen Staab; Moritz Hardt; Maksym Andriushchenko; Jonas Geiping; |
| 74 | AgentVista: Evaluating Multimodal Agent in Ultra-Challenging Realistic Visual Scenarios Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce AgentVista, a benchmark for generalist multimodal agents that spans 25 sub-domains across 7 categories, pairing realistic and detail-rich visual scenarios with natural hybrid tool use. |
Zhaochen Su; Jincheng Gao; Hangyu Guo; Zhenhua Liu; Lueyang Zhang; Xinyu Geng; Shijue Huang; Peng Xia; Guanyu Jiang; Cheng Wang; Yue Zhang; Yi R. (May) Fung; Junxian He; |
| 75 | Negation Neglect: When Models Fail to Learn Negations in Training Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce **Negation Neglect**, a phenomenon where finetuning LLMs on documents that flag a claim as false leads them to believe the claim is true. |
Harry Mayne; Lev McKinney; Jan Dubiński; Adam Karvonen; James Chua; Owain Evans; |
| 76 | Visual-ERM: Reward Modeling for Visual Equivalence Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Visual Equivalence Reward Model (Visual-ERM), a multimodal generative reward model that provides fine-grained, interpretable, and task-agnostic feedback to evaluate vision-to-code quality directly in the rendered visual space. |
Ziyu Liu; Shengyuan Ding; Xinyu Fang; Xuanlang Dai; Penghui Yang; Jiaqi Wang; Kai Chen; Dahua Lin; Yuhang Zang; |
| 77 | Relaxed On-Policy Distillation: Selective Credit Allocation for Scaling Reasoning Efficiently Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Most tokens carry redundant near-zero credit, while rare heavy-tailed negative credits can dominate updates and prematurely suppress plausible reasoning trajectories. We introduce REOPOLD (Relaxed On-Policy Distillation), a framework that relaxes this uniform assignment by controlling where teacher feedback is applied, how strongly it affects the update, and when the allocation rule changes during training. |
Jongwoo Ko; Sara Abdali; Young Jin Kim; Tianyi Chen; Pashmina Cameron; |
| 78 | Diffusion Fine-Tuning: Iterative Refinement for Advanced Grounding with Diffusion Large Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We attribute this bottleneck to two properties of the vertex-level autoregressive (AR) paradigm used by current LVLMs: (1) errors in early-emitted vertices propagate uncorrected through the rest of the sequence, and (2) the model commits to local vertex placement before observing the full contour, leading to suboptimal allocation of a fixed vertex budget. We propose Diffusion Fine-Tuning (DFT), which removes the vertex-level sequential dependency by placing all 2N coordinate tokens under a single discrete-diffusion denoiser, while introducing a much shorter digit-level factorisation: each coordinate is decomposed into hundreds, tens, and units digits, and the reverse process is trained to predict them coarse-to-fine. |
Zhangyang Qi; Jinsong Li; Jiaqi Wang; Hengshuang Zhao; |
| 79 | LEAP: Library-driven Evolutionary Abstraction Paradigm for Large Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: A fundamental limitation of current evolutionary code generation is its reliance on unguided stochasticity to derive child programs from parents, lacking a systematic mechanism for knowledge accumulation, whereas human problem-solving thrives on the continuous consolidation and reuse of high-level conceptual abstractions. To bridge this gap, we propose the \textbf{Library-driven Evolutionary Abstraction Paradigm (LEAP)}, a cognitive-inspired framework that enables LLMs to autonomously construct and evolve a globally shared library of reusable algorithmic primitives. |
Fanbin Lu; Chi-Wing Fu; Jiaya Jia; |
| 80 | FASTER: Value-Guided Sampling for Fast RL Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose FASTER, a method for getting the benefits of sampling-based test-time scaling of diffusion-based policies without the computational cost by tracing the performance gain of action samples back to earlier in the denoising process. |
Perry Dong; Alexander Swerdlow; Dorsa Sadigh; Chelsea Finn; |
| 81 | ViCO: A Training Strategy Towards Semantic Aware Dynamic High-Resolution Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose Visual Consistency Learning (ViCO), a novel training algorithm that enables the model to represent images of varying semantic complexities using different numbers of vision tokens. |
Long Cui; Weiyun Wang; Jie Shao; Zichen Wen; Gen Luo; Zhang; Linfeng Zhang; Wenhai Wang; |
| 82 | Network of Theseus (Like The Ship) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: A standard assumption in deep learning is that the inductive bias introduced by a neural network architecture must persist from training through inference. The architecture you … |
Vighnesh Subramaniam; Colin Conwell; Boris Katz; Andrei Barbu; Brian Cheung; |
| 83 | The Optimization Prior: Instilling Depth for Shallow Networks, Detail for Coarse Networks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Because this starting point dictates the model’s future trajectory, we ask: can this initial geometry be explicitly chosen by referencing a completely different architecture? To achieve this, we introduce representational similarity as an optimization prior. |
Vighnesh Subramaniam; Boris Katz; Brian Cheung; |
| 84 | ODEWorld: A Continuous Predictive Architecture Via Physical-Time Flow Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, existing machine learning paradigms for world modeling are largely confined to discrete-time prediction and inference, thereby exhibiting significant inefficiency in capturing the fundamental dynamics of the physical world. To bridge this gap, we introduce Physical-Time Flow (**PT-Flow**), a novel approach that learns a continuous latent velocity field operating in physical time. |
Dongxiu Liu; Haoyi Niu; Peng Cheng; Yuan Gao; Xirui Kang; Sangli Teng; Koushil Sreenath; Xianyuan Zhan; |
| 85 | You Only Need Minimal RLVR Training: Extrapolating LLMs Via Rank-1 Trajectories Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we demonstrate that RLVR weight trajectories are extremely low-rank and highly predictable. |
Zhepei Wei; Xinyu Zhu; Wei-Lin Chen; Chengsong Huang; Jiaxin Huang; Yu Meng; |
| 86 | Unveiling Fine-Grained Visual Traces: Evaluating MultiModal Interleaved Reasoning Chains in Multimodal STEM Tasks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: STEM reasoning is a particularly valuable testbed because it provides highly verifiable feedback, but existing benchmarks often permit unimodal shortcuts due to modality redundancy and focus mainly on final-answer accuracy, overlooking the reasoning process itself. To address this challenge, we introduce STEPSTEM: a graduate-level benchmark of 283 problems across mathematics, physics, chemistry, biology, and engineering for fine-grained evaluation of cross-modal reasoning in MLLMs. |
Jing Jin; Hao Liu; Yan Bai; Yihang Lou; Zhenke Wang; Tianrun Yuan; Yongkang Zhu; Juntong Chen; Fanhu Zeng; Xuanyu Zhu; Tao Feng; Yige Xu; |
| 87 | BBOWP-Bench: Evaluating LLMs on Black-Box Optimization Word Problems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This paper introduces Black-Box Optimization Word Problems (BBOWP), a novel problem setting in which a system must infer both a search space and an optimization algorithm from a natural-language description of a black-box optimization task. |
Yutaro Yamada; Kei Hiroshima; Nozomu Yoshinari; Kento Uchida; Shinichi Shirakawa; |
| 88 | A Process-Level Evaluation of LLM Discovery Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we introduce a four-level ladder of agent harnesses and a process-level evaluation framework that measures behavioral profiles and intermediate milestones, alongside the final performance. |
Haorui Wang; Yuanqi Du; Hao Cheng; Baolin Peng; Chao Zhang; Weizhu Chen; yelong shen; |
| 89 | Learning Process Rewards Via Visitation Matching for Efficient RL Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a simple approach to transform a sparse outcome reward into a dense process reward. |
Raymond Tsao; Andrew Wagenmaker; Sergey Levine; |
| 90 | Test-Time Gradient Guidance of Flow Policies in Reinforcement Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: It often requires specialized training objectives or back-propagating through denoising processes, which cause well known issues with stability and affects scalability. In this paper we study the question of whether simple policy improvement schemes at test time alone, leaving stable supervised policy training intact, can be a competitive alternative which side-steps these issues. |
Zhiyuan (Paul) Zhou; Andy Peng; Charles Xu; Qiyang Li; Jost Springenberg; Kevin Frans; Sergey Levine; |
| 91 | SLVR: Structured Latent Visual Reasoning Via Human-like Reasoning Flows Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Structured Latent Visual Reasoning (SLVR), a training framework that bridges explicit chain-of-thought and latent reasoning by organizing multimodal reasoning into typed latent stages for planning, grounding, evidence selection, and reasoning integration. |
Albert Gao; Bing Xue; Andrea Zanette; |
| 92 | ECHO: Terminal Agents Learn World Models for Free Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce ECHO (Environment Cross-entropy Hybrid Objective), an auxiliary loss that trains the policy to predict observation tokens alongside the policy gradient. |
Vaishnavi Shrivastava; Ahmed Awadallah; Dimitris Papailiopoulos; |
| 93 | Dr. MAS: Stable Reinforcement Learning for Multi-Agent LLM Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We theoretically pinpoint a key source of instability when extending group-based RL to cooperative multi-agent LLM systems: under GRPO-style optimization, a global normalization baseline can mismatch heterogeneous agents’ reward distributions, inducing gradient-norm instability. Based on this finding, we propose Dr. MAS, a simple and stable RL recipe for cooperative multi-agent LLM systems. |
Lang Feng; Longtao Zheng; Shuo He; Fuxiang Zhang; Bo An; |
| 94 | BalCapRL : A Balanced Framework for RL-Based MLLM Image Captioning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: For example, utility-oriented objectives can encourage noisy, hallucinated, or overlong captions that improve downstream question answering while harming fluency, whereas arena-style objectives can favor fluent but generic descriptions with limited usefulness. To address this, we propose a more balanced RL framework that jointly optimizes utility-aware correctness, reference coverage, and linguistic quality. |
Shaokai Ye; Vasileios Saveris; Yihao Qian; Jiaming Hu; Elmira Amirloo; Peter Grasch; |
| 95 | Generative Modeling Via Drifting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose *Drifting Models*, a generative modeling framework that evolves the pushforward distribution during training and naturally admits one-step inference. |
Mingyang Deng; He Li; Tianhong Li; Yilun Du; Kaiming He; |
| 96 | Training Optimal Large Diffusion Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Quokka, the first comprehensive scaling law for diffusion language models (DLMs), encompassing both compute- and data-constrained regimes alongside key modeling and optimization designs. |
Jinjie Ni; Qian Liu; Chao Du; Longxu Dou; Hang Yan; Zili Wang; Tianyu Pang; Michael Shieh; |
| 97 | Holo4D: Holistic 4D Reconstruction As Geometric Control for Video Diffusion Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce HOLO4D, a geometry-aware video-to-video framework that controls a pretrained VDM with a hybrid 4D geometric cache built from the source video. |
Yushi LAN; Zeren Jiang; Koichi Namekata; Xingang Pan; Chuanxia Zheng; Andrea Vedaldi; |
| 98 | Track4D: Representing Dense 3D Tracking for Video Diffusion Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Track4D, a method that formulates dense 3D tracking as conditional video generation. |
Yushi LAN; Zeren Jiang; Kelvin Zheng Li; Xingang Pan; Chuanxia Zheng; Andrea Vedaldi; |
| 99 | Fairness Failure Modes of Multimodal LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose two metrics Fairness Score and Bias Score and design an evaluation paradigm to address shortcut reasoning and data contamination challenges.We release our code and dataset to facilitate further evaluations and the development of mitigation methods here. |
Canyu Chen; Anglin Cai; Joan Nwatu; Yale Li; Jessica Hullman; Rada Mihalcea; Kathleen McKeown; Manling Li; |
| 100 | SCULPT: Advancing Masked Discrete Diffusion for High-Resolution Image Synthesis Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: we proposed a Goruped Corss Entrophy (GCE) objective that assigns positive learning signal to adjacent tokens of the ground truth in the embedding space. |
Shufan Li; Greg Heinrich; Hanrong Ye; Yonggan Fu; Aditya Grover; Jan Kautz; Pavlo Molchanov; |
| 101 | EcoGym: Evaluating LLMs for Long-Horizon Plan-and-Execute in Interactive Economies Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce EcoGym, a generalizable benchmark for continuous plan-and-execute decision making in interactive economies. |
Xueyu Hu; Jinxiang Xia; Shengze Xu; Kangqi Song; Yishuo Yuan; Guibin Zhang; JinCheng Ren; Boyu Feng; Li Lu; Tieyong Zeng; Jiaheng Liu; Minghao Liu; He Zhu; Eleanor Jiang; Wei Wang; Wangchunshu Zhou; |
| 102 | Higher-Order Action Supervision Makes A Strong Policy Class Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we show that simultaneously supervising both zeroth- and first-order actions can dramatically enhance policies’ performance and control robustness. |
Peng Cheng; Yunxian Hou; Zhi Zhou; Qian Zhang; Chang Huang; Xianyuan Zhan; |
| 103 | Multiclass Classification with Rare Useful Features: Fundamental Limits and The Optimality of Diversity Pursuit Higher Criticism Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We focus on a rare feature diversity model, in which most features are useless, but a small fraction promote discrimination among classes in a way that can be detected by a per-feature diversity test, such as one-way ANOVA or the Kruskal–Wallis test. For this setting, we propose Diversity Pursuit Higher Criticism (DP-HC), a feature-screening procedure that applies a higher-criticism threshold to the collection of p-values obtained from per-feature diversity tests. |
Gabriel Moryoussef; Alon Kipnis; |
| 104 | Masked Diffusion Language Models Are Strong and Steerable Text-Based World Models for Agentic RL Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we (i) formalize text-based world modeling as a steerable transition-dynamics problem decomposed into initial environment state, task context, tool schemas, domain rules, and steering directives, and (ii) curate a dataset of 239,403 grounded state–action trajectories spanning nine open-source environments and twelve frontier model families. |
Darshan Deshpande; |
| 105 | CausalSpatial: A Benchmark for Object-Centric Causal Spatial Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our analysis uncovers a fundamental deficiency: models over-rely on textual chain-of-thought reasoning that drifts from visual evidence, producing fluent but spatially ungrounded hallucinations. To address this, we propose the Causal Object World Model (COW), a framework that externalizes the simulation process by generating videos of hypothetical dynamics. |
Wenxin (Wendy) Ma; Chenlong Wang; Ruisheng Yuan; Hao Chen; Nanru Dai; Chengxin Qian; Yijun Yang; Qi Chen; Zhaoyang Wang; S. Kevin Zhou; Jianwen Xie; Alan Yuille; Jieneng Chen; |
| 106 | Tango3D: Towards Alignment for Global and Local 2D-3D Correspondence Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, this global-only alignment cannot establish fine-grained pixel-to-point correspondence. To solve this, we present Tango3D, a foundation model that unifies dense correspondence and global retrieval. |
Zebin He; Mingxin Yang; Shuhui Yang; Hanxiao Sun; Xintong Han; Chunchao Guo; Wenhan Luo; |
| 107 | When to Adopt Model Updates Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce a framework for assessing when to propagate newly released model versions to downstream applications that does not require access to upstream data or retraining a priori. |
Aspen Hopkins; Isabella Struckman; Hedi Driss; Aleksander Madry; |
| 108 | One Language-Free Foundation Model Is Enough for Universal Vision Anomaly Detection Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we rethink the fundamental mechanism of vision-language models for AD and present UniADet, an embarrassingly simple, effective and general framework for Universal vision Anomaly Detection. |
Bin-Bin Gao; |
| 109 | Conditional Misalignment: Common Interventions Can Hide Emergent Misalignment Behind Contextual Triggers Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Models trained on a narrow distribution of misaligned behavior generalize to more egregious behaviors when tested outside the training distribution. |
Jan Dubiński; Jan Betley; Anna Sztyber-Betley; Daniel Tan; Owain Evans; |
| 110 | Fev-bench: A Realistic Benchmark for Time Series Forecasting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Many benchmarks lack consistent evaluation infrastructure or are too rigid for integration into existing pipelines. To address these gaps, we propose fev-bench, a benchmark of 100 forecasting tasks across seven domains, including 46 with covariates. |
Oleksandr Shchur; Abdul Fatir Ansari; Ali Caner Turkmen; Lorenzo Stella; Nick Erickson; Pablo A Guerron; Michael Bohlke-Schneider; Yuyang (Bernie) Wang; |
| 111 | HandEdit: A Unified Benchmark for Egocentric Human-to-Robot Dexterous Hand Image Editing Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we present HandEdit, a unified large-scale embodiment-aware image-editing dataset and benchmark specifically designed to transform human hands and arms into various dexterous robotic embodiments within egocentric frames. |
Zhenjie Yang; Xingyu Jiao; Guopeng Zhong; Shuzhe Yang; Shi Che; Chao Wu; Chenyu Jiang; Dongjie Zhang; Yideng Zhang; Zheng Zhang; Muyun Jiang; Haisheng Su; Shuang Jin; Donghang Zhang; Chao Yang; Li Chen; Hongyang Li; Zuxuan Wu; Junchi Yan; Xiaosong Jia; Yu-Gang Jiang; |
| 112 | General Agent Evaluation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To enable this study, we contribute (1) a unifying protocol that bridges existing benchmark and agent protocols; (2) an evaluation harness that surfaces any benchmark to any general-purpose agent and backbone model; and (3) the first Open General Agent Leaderboard of agent configurations, a full factorial over 5 agent architectures × 5 backbone LLMs (three closed-source, two open-weight) × 6 benchmarks spanning software engineering, customer service, deep research, and personal assistance. |
Elron Bandel; Asaf Yehudai; Lilach Edelstein; Yehoshua Sagron; Yotam Perlitz; Elad Venezian; Natalia Razinkov; Natan Ergas; Shlomit S Ifergan; Segev Shlomov; Michal Jacovi; Leshem Choshen; Liat Ein-Dor; Yoav Katz; Michal Shmueli-Scheuer; |
| 113 | On The Selectivity of Generative Models in Structure-Based Drug Design Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Although structure-based drug design (SBDD) has advanced rapidly with deep learning, most generative models still optimize binding to a single target and disregard selectivity due to limited training data and unknown off-target interactions. Motivated by this gap, we introduce SelectBench, a comprehensive benchmark assessing the selectivity of generative SBDD models. |
Ella Miray Rajaonson; Jungyoon Lee; William Chau; Alan Aspuru-Guzik; Benjamin Sanchez-Lengeling; Dominique Beaini; Kirill Neklyudov; Marta Skreta; |
| 114 | Long-Horizon Q-Learning: Accurate Value Learning Via N-Step Inequalities Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose *long-horizon Q-learning* (LQL), which introduces a principled backstop against compounding error when learning the optimal action-value function. |
Armaan A Abraham; Lucy Xiaoyang Shi; Chelsea Finn; |
| 115 | Castle-in-the-Air: Probing The Foundational Visual Deficits of MLLMs Via Bottom-Up Cognitive Factors Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Does it guarantee their performance on these foundational visual capabilities? To systematically investigate this gap, we introduce VisFactor, a benchmark that digitizes 20 vision-centric subtests from FRCT, a well-established cognitive psychology assessment spanning four domains of human visual cognition. |
Jen-Tse Huang; Dasen Dai; Jen-Yuan Huang; Youliang Yuan; Xiaoyuan Liu; Wenxuan Wang; Wenxiang Jiao; Pinjia He; Zhaopeng Tu; Haodong Duan; |
| 116 | FiedlerPrune: Connectivity-Preserving Cross-Layer Pruning for Large Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose FiedlerPrune, a pruning framework that achieves a better balance between the two paradigms by incorporating cross-layer dependencies. |
Zijun Sun; Yanning Shen; |
| 117 | Action Images: End-to-End Policy Learning Via Multiview Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we present Action Images, a unified world action model that formulates policy learning as multiview video generation. |
Haoyu Zhen; Zixian Gao; Qiao Sun; yilin zhao; Yuncong Yang; Yilun Du; Pengsheng Guo; Tsun-Hsuan Johnson Wang; Yi-Ling Qiao; Chuang Gan; |
| 118 | Trajectory-Consistent Diffusion Policies for Offline Reinforcement Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This issue is particularly pronounced on mixed-quality offline datasets, where suboptimal behaviors amplify trajectory instability. To address this problem, we propose Trajectory Consistency Rectification (TCR), a framework designed to promote denoising stability during both training and inference. |
Yichao Fu; Shangde Gao; Zhuoling Li; Wen Wang; Shangqi Gao; Ke Liu; |
| 119 | Recursive Multi-Agent Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To this end, we introduce RecursiveMAS, a recursive multi-agent framework that casts the entire system as a unified latent-space recursive computation. |
Jiaru Zou; Rui Pan; Ruizhong Qiu; Pan Lu; Shizhe Diao; Jindong Jiang; Hanghang Tong; Tong Zhang; Markus Buehler; Jingrui He; James Zou; |
| 120 | Grid Games: The Power Of Multiple Grids for Quantizing Large Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we study the following natural extension: assume that, for each group of values, we are free to select the better among two or more 4-bit grids marked by one or more bits in the scale value. |
Vage Egiazarian; Erik Schultheis; Andrei Panferov; Earl Killian; Torsten Hoefler; Dan Alistarh; |
| 121 | The Era of Agentic Organization: Learning to Organize with Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Specifically, we propose a thinking protocol where an organizer dynamically assigns sub-queries to workers, merges intermediate knowledge, and produces coherent solutions. |
Zewen Chi; Li Dong; Qingxiu Dong; Yaru Hao; Xun Wu; Shaohan Huang; Furu Wei; |
| 122 | On The Geometry and Latent-Space Composition of Hypernetwork-Generated LoRAs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we study the problem of composing hypernetwork-generated LoRAs using two representative systems, Doc-to-LoRA and SHINE. |
Yanlai Yang; Minki Kang; Divyam Madaan; Sung Ju Hwang; Mengye Ren; |
| 123 | Evolving Agent Teams Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, we find that existing coding agents converge to local optima on real-world, expert-written kernels (e.g., FlashInfer). We introduce Evolving Agent Teams, a multi-agent framework that escapes such local optima. |
Shiyi Cao; Ziming Mao; Dacheng Li; Joseph Gonzalez; Ion Stoica; |
| 124 | ReGuidance: Diffusion Steering with Strong Latent Initializations Solves Hard Inverse Problems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Training-free methods like gradient guidance have offered simple, flexible approaches for these tasks, but when the reward is not informative enough, e.g., in inverse problems with highly compressive measurements, these techniques can veer off the data manifold, failing to produce realistic data samples. To address this challenge, we devise a simple algorithm, *ReGuidance*, that leverages prior methods’ solutions as strong initializations and substantially enhancing their realism. |
Aayush Karan; Kulin Shah; Sitan Chen; |
| 125 | VUM: Visual Unified Models for Image Generation and Perception Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Thus, existing works rely on decoupled architectures, as forcing these opposing information flows into one network inevitably triggers task interference and representation collapse. To break this bottleneck, we introduce the Visual Unified Model (VUM), a native framework that unifies both capabilities within a single architecture. |
ZiDong Wang; Yiyuan Zhang; Xiaoyu Yue; Xiangyuan Xue; Zhangquan Chen; Manyuan Zhang; Wanli Ouyang; |
| 126 | DiT4DiT: Jointly Modeling Video Dynamics and Actions for Generalizable Robot Control Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Generative video models, by contrast, already encode motion, contact, and implicit physics at internet scale. We introduce DiT4DiT, a Video-Action Model (VAM) that couples a Video Diffusion Transformer with an Action Diffusion Transformer and co-trains them under a dual flow-matching objective with a tri-timestep scheme for video supervision, action supervision, and feature extraction. |
Teli Ma; Jia Zheng; Zifan Wang; Chunli Jiang; Andy Cui; Junwei Liang; Shuo Yang; |
| 127 | RAVEN: Real-time Autoregressive Video Extrapolation with Consistency-model GRPO Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce the Real-time Autoregressive Video Extrapolation Network (RAVEN), a training-time test framework that repacks each self rollout into an interleaved sequence of clean historical endpoints and noisy denoising states. |
Yanzuo Lu; Ronglai Zuo; Jiankang Deng; |
| 128 | Don’t Discard Your Rollouts: Reusing Teacher RL Traces for Student Distillation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we show that rollout reuse becomes effective when it is treated as a *data-selection problem. |
Vashisth Tiwari; Emma Strubell; Zico Kolter; |
| 129 | D-OPSD: On-Policy Self-Distillation for Continuously Tuning Step-Distilled Diffusion Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: For example, applying the commonly used fine-tuning technique would compromise their inherent few-step inference capability. To address this, we propose D-OPSD, a novel training paradigm for step-distilled diffusion models that enables on-policy learning during supervised fine-tuning. |
Dengyang Jiang; Xin Jin; Dongyang Liu; Zanyi Wang; Mingzhe Zheng; Ruoyi Du; Xiangpeng Yang; Qilong Wu; Zhen Li; Peng Gao; Harry Yang; Steven Hoi; |
| 130 | Autonomous Continual Learning for Environment Adaptation of Computer-Use Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce **ACuRL**, an **A**utonomous **Cu**rriculum **R**einforcement **L**earning framework that continually adapts agents to specific environments with zero human data. |
Tianci Xue; Zeyi Liao; Tianneng Shi; Zilu Wang; Kai Zhang; Dawn Song; Yu Su; Huan Sun; |
| 131 | Beyond Distribution Matching: Self-Supervised Representation Forcing for Few-Step Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: As the step count shrinks, this output-only supervision becomes too coarse, and the distilled student degrades on both fine spatial structure and coherent temporal dynamics. To close this gap, we introduce Representation-Forcing, which adds a predictive representation loss on top of DMD without changing its output objective. |
Junyi Chen; Haiyu Zhang; ZHOUJIE FU; Tengfei Wang; Chunchao Guo; |
| 132 | Training Deliberative Monitors for Black-Box Scheming Detection Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we study action-only deliberative monitors: smaller open-weight models trained to detect scheming and sabotage from agentic trajectories without accessing the monitored agent’s reasoning or model internals. |
Aditya Sinha; Akshat Naik; Victor Gillioz; Simon Storf; Kilian Merkelbach; Rich Barton-Cooper; Axel Højmark; Marius Hobbhahn; |
| 133 | Medmarks: A Comprehensive Open-Source LLM Benchmark Suite for Medical Tasks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Medmarks, a fully open-source evaluation suite with 30 benchmarks spanning question answering, information extraction, medical calculations, and open-ended clinical reasoning. |
Benjamin Warner; Ratna S Grandhi; Max Kieffer; Aymane Ouraq; Saurav Panigrahi; Geetu Ambwani; Kunal Bagga; Nikhil Khandekar; Arya Hariharan; Nishant Mishra; Manish Saravanan; Shamus S Yang; Ahmed Essouaied; Adepoju J Moyondafoluwa; Robert Scholz; Bofeng Huang; Molly Beavers; Srishti Gureja; Anish Mahishi; Sameed Khan; Maxime Griot; Hunar Batra; Jean-Benoit Delbrouck; Siddhant Bharadwaj; Ronald Clark; Ashish Vashist; Anas Zafar; Leema K Murali; Harsh Deshpande; Ameen Patel; William Brown; Johannes Hagemann; Connor Lane; Paul Scotti; Tanishq M Abraham; |
| 134 | WordEval: Evaluating Word-Native Operation Fidelity in Document Editing Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce WordEval, a benchmark for evaluating Word-native document editing from natural-language instructions. |
Jinyoung Park; Yupan Huang; Shimin Zhang; Jiayu Ding; Yilin Jia; Yuzhong Zhao; Tengchao Lv; Wenshan Wu; Shaohan Huang; Nan Yang; Xiangyang Zhou; Dongdong Zhang; Li Dong; Lei Cui; Sun Mao; Xun Wang; Tao Ge; Huitian Jiao; Si-Qing Chen; Changick Kim; Furu Wei; |
| 135 | InfiniteVL: A Systematic Approach to Highly-Efficient, Ultra-Long Multimodal Understanding Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This systematic challenge is essentially three-fold: (1) \textbf{Efficient Architecture}: compressing redundant visual sequences without sacrificing fine-grained details; (2) \textbf{Knowledge Transfer}: seamlessly migrating the capabilities of strong pretrained VLMs into more efficient novel architectures; and (3) \textbf{Generalization}: maintaining robust perception and reasoning across diverse multimodal domains. To address this, we introduce \textbf{InfiniteVL}, a linear-sparse hybrid VLM framework designed for efficient long-context understanding. |
Hongyuan Tao; Bencheng Liao; Shaoyu Chen; haoran yin; Qian Zhang; Wenyu Liu; Xinggang Wang; |
| 136 | Learning Transferable Representations from Operating System Entities Via Provenance Graph Distillation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: These behavioral regularities can be learned offline from provenance graphs and approximated at inference time from entity labels alone. Based on this insight, we introduce SPIDER (System Provenance-Informed Distilled Entity Representations), a text-only entity encoder distilled from provenance graphs. |
Tristan Bilot; Xueyuan Han; Thomas Pasquier; |
| 137 | Tcell: Mitigating Harmful Fine-tuning for Large Language Models Via Gradient Alignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we first explore an intuitive gradient mixing solution, and derive a key property ensuring the success of defense — \emph{taking a fine-tuning update that has higher cosine similarity with the safety gradient can mitigate harmful fine-tuning.} Motivated by this key property, we design an alignment-stage defense, dubbed Tcell. |
Tiansheng Huang; Virat Vishnu Shejwalkar; Oscar Chang; Ling Liu; |
| 138 | Chain-of-Thought Is Not Explainability Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We analyze *faithfulness* as whether CoTs are not only human-interpretable, but also reflect the model’s underlying reasoning well enough to support responsible use. |
Fazl Barez; Tung-Yu Wu; Iván Arcuschin Moreno; Michael Lan; Vincent Wang-Maścianica; Noah Siegel; Nicolas Collignon; Clement Neo; Isabelle Lee; Alasdair Paren; Adel Bibi; Robert Trager; Damiano Fornasiere; John Yan; Yanai Elazar; Yoshua Bengio; |
| 139 | Self-Distilled RLVR Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Accordingly, we identify the optimal niche for self-distillation and propose \textbf{RLSD} (\textbf{RL}VR with \textbf{S}elf-\textbf{D}istillation). |
Chenxu Yang; Chuanyu Qin; Qingyi Si; Minghui Chen; Naibin Gu; Dingyu Yao; Zheng Lin; Weiping Wang; Nan Duan; Jiaqi Wang; |
| 140 | DMoE: DLLMs with Learnable Block Experts Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This mismatch substantially increases the number of uniquely activated experts, making inference increasingly memory-bound. To address this, we propose dMoE, a simple yet effective block-level MoE framework. |
Sicheng Feng; Zigeng Chen; Gongfan Fang; Xinyin Ma; Xinchao Wang; |
| 141 | Unified Generative-Predictive Modeling for 4D Scene Understanding Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a unified generative-predictive framework for 4D scene understanding that jointly models RGB video and geometric representations such as depth and camera pose with a single joint generative model. |
Amani Kiruga; Zhiyi Li; Ruojin Cai; Hansen Lillemark; Xiaoshan Wu; Ayush Tewari; Yilun Du; |
| 142 | The Scaling Laws of Skills in LLM Agent Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Across 15 frontier LLMs, 1,141 real-world skills, and over 3M routing or execution decisions, we identify two coupled laws. |
Qiguang Chen; Qiming Yu; Yuhang Gu; Zhuoye Huang; Hanjing Li; Hongyu Liu; Simin Liu; Jinhao Liu; Dengyun Peng; Jiangyi Wang; Zheng Yan; Fanqing Meng; Ethan Qin; Carl Che; Mengkang Hu; |
| 143 | Co-GRPO: Co-Optimized Group Relative Policy Optimization for Masked Image Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we introduce Co-GRPO, which reformulates MDM generation as a unified Markov Decision Process (MDP) that jointly incorporates both the model and the inference schedule. |
Renping Zhou; Zanlin Ni; Tianyi Chen; Zeyu Liu; Yang Yue; Yulin Wang; Yuxuan Wang; Jingshu Liu; Gao Huang; |
| 144 | Liars’ Bench: Evaluating Lie Detectors for Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Prior work has introduced techniques for detecting when large language models (LLMs) lie, that is, generate statements they believe are false. |
Kieron Kretschmar; Walter Laurito; Sharan Maiya; Samuel Marks; |
| 145 | Data Diversity Drives The Emergence of Symbolic Mechanisms Supporting Abstract Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Recent work has identified a set of emergent symbolic mechanisms that support abstract reasoning in large language models, but it remains unclear what factors drive the emergence of these mechanisms. To address this question, we trained neural networks from scratch on abstract sequence tasks, varying both architectural and data distributional factors, and investigated the effects of these factors on the emergence of symbolic mechanisms. |
Melody Li; Taylor Webb; |
| 146 | Speculative Self-Distillation Enables Efficient Knowledge Internalization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Speculative Self-Distillation (SSD), an efficient per-token mixed-policy method for self-distillation. |
Shayan Talaei; Agam Bhatia; Arshia Soltani Moakhar; Jonas Hübotter; Amin Saberi; Azalia Mirhoseini; |
| 147 | Finetuning with Sampling: Make SFT Generalize, Not Forget Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To this end, we introduce a Markov chain Monte Carlo (MCMC) sampling algorithm that transforms off-policy traces into more on-policy ones given a reference model for finetuning. |
Aayush Karan; Sitan Chen; Yilun Du; |
| 148 | ProgramBench: Can Language Models Rebuild Programs From Scratch? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, existing benchmarks measure focused, limited tasks such as fixing a single bug or developing a single, specified feature. We therefore introduce ProgramBench to measure the ability of software engineering agents to develop software holisitically. |
John Yang; Kilian Lieret; Jeffrey Ma; Parth Thakkar; Dmitrii Pedchenko; Sten Sootla; Emily McMilin; Pengcheng Yin; Rui Hou; Gabriel Synnaeve; Diyi Yang; Ofir Press; |
| 149 | On Data Engineering for Scaling LLM Terminal Capabilities Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Despite rapid recent progress in the terminal capabilities of large language models, the training data strategies behind state-of-the-art terminal agents remain largely undisclosed. We address this gap through a systematic study of data engineering practices for terminal agents, making two key contributions: (1) Terminal-Task-Gen, a lightweight synthetic task generation pipeline that supports seed-based and skill-based task construction, and (2) a comprehensive analysis of data and training strategies, including filtering, curriculum learning, long context training, and scaling behavior. |
Renjie Pi; Grace Lam; Mohammad Shoeybi; Pooya Jannaty; Bryan Catanzaro; Wei Ping; |
| 150 | Reward-Conditioned Reinforcement Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Reward-Conditioned Reinforcement Learning (RCRL), an off-policy method that conditions agents on reward parameterizations while collecting experience under a single nominal objective. |
Michal Nauman; Marek Cygan; Pieter Abbeel; |
| 151 | MeshFIM: Local Low-Poly Mesh Editing Via Fill-in-the-Middle Autoregressive Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce MeshFIM, a Fill-in-the-Middle (FIM) framework that regenerates a target region of a low-poly mesh conditioned on the surrounding context. |
Dingdong Yang; Jian Liu; Biwen Lei; Haohan Weng; Zhuo Chen; Song Guo; Hao Zhang; Ali Mahdavi Amiri; Chunchao Guo; |
| 152 | Next Forcing: Causal World Modeling with Multi-Chunk Prediction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we present Next Forcing, a multi-chunk prediction (MCP) framework for causal world modeling that enables faster training, higher accuracy, and accelerated inference. |
Gangwei Xu; Qihang Zhang; Jiaming Zhou; Xing Zhu; Yujun Shen; Xin Yang; Yinghao Xu; |
| 153 | LaST-VLA: Thinking in Latent Spatio-Temporal Space for Vision-Language-Action in Autonomous Driving Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, without explicit intermediate constraints, standard latent CoT often operates as a physics-agnostic representation. To address this, we propose the **La**tent **S**patio-**T**emporal **VLA** (**LaST-VLA**), a framework shifting the reasoning paradigm from discrete symbolic processing into a physically grounded Latent Spatio-Temporal CoT. |
Yuechen Luo; Fang Li; Shaoqing Xu; Yang Ji; Zehan Zhang; Bing Wang; Shen Yuannan; Jianwei Cui; Long Chen; Guang Chen; Hangjun Ye; Zhi-Xin Yang; Fuxi Wen; |
| 154 | Flash-KMeans: Fast and Memory-Efficient Exact K-Means Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: k-means has historically been positioned primarily as an offline processing primitive, typically used for dataset organization or embedding preprocessing rather than as a first-class component in online systems. In this work, we revisit this classical algorithm under the lens of modern AI system design and enable k-means as an online primitive. |
Shuo Yang; Haocheng Xi; Yilong Zhao; Muyang Li; Xiaoze Fan; Jintao Zhang; Han Cai; Yujun Lin; Xiuyu Li; Kurt Keutzer; Song Han; Chenfeng Xu; Ion Stoica; |
| 155 | Steering Visual Generation in Unified Multimodal Models with Understanding Supervision Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This design, while effective for individual tasks, weakens the connection required for mutual enhancement, leaving the potential synergy empirically uncertain. We propose to explicitly restore this synergy by introducing ***Un**derstanding-**O**}riented Post-Training* (**UNO**), a lightweight framework that treats understanding not only as a distinct task, but also a direct supervisory signal to steer generative representations. |
Zeyu Liu; Zanlin Ni; Yang Yue; Cheng Da; Huan Yang; Di ZHANG; Kun Gai; Gao Huang; |
| 156 | Tyche: One Step Flow for Efficient Probabilistic Weather Forecasting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, their inference cost scales with forecast horizon, ensemble size, and the number of denoising steps required for each transition, making large operational ensembles expensive. To address this, we present Tyche, a one-step conditional flow model for efficient probabilistic weather forecasting. |
Fan Xu; Yuan Gao; Kun Wang; Rui Su; Fenghua Ling; Hao Wu; Wanli Ouyang; |
| 157 | Behaviour4All: A Dependency-Aware Toolkit for In-the-wild Facial Behaviour Analysis Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Behaviour4All, the first dependency-aware toolkit that unifies VA, EXP, AUs via explicit relational priors and principled optimisation framework. |
Dimitrios Kollias; Chunchang Shao; Odysseus Kaloidas; Ioannis Patras; |
| 158 | Programmatic Reasoning with Structural Schema: A Unified Framework for Multi-Table Inference Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce StrucTab-R1, a schema-centric reasoning framework that decouples relational reasoning from raw data exposure. |
Jialin Chen; Brandon Mayer; Michael Galkin; Sami Abu-El-Haija; Rex Ying; Bryan Perozzi; |
| 159 | Latent Introspection: Models Can Detect Prior Concept Injections Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We uncover a latent capacity for introspection in a Qwen 32B model, demonstrating that the model can detect when concepts have been injected into its earlier context and identify which concept was injected. |
Theia Pearson-Vogel; Martin Vanek; Raymond Douglas; Jan Kulveit; |
| 160 | Characterizing The Aesthetic Defaults of Generative Image Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce LouvreSAE, a sparse autoencoder over CLIP image embeddings whose dictionary is shaped to fall on stylistic rather than object features, and LatentAesthetic, an evaluation methodology that uses it to estimate, ground, and compare generators’ default aesthetics under content-controlled, style-neutral prompts. |
Maty Bohacek; Raina Panda; Daniel Fein; Arpita Singhal; Mark Fiore; Bill Freeman; Maneesh Agrawala; |
| 161 | SCALECUA : Scaling Computer Use Agents with Verifiable Task Synthesis and Efficient Online RL Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, applying this paradigm is severely bottlenecked by verifiable data scarcity and online RL inefficiency. To break these barriers, we introduce **ScaleCUA**, a unified framework that scales online RL for CUAs by combining verifiable task synthesis with efficient online RL. |
Bowen Lv; Xiao Liu; Yanyu Ren; Hanyu Lai; Bohao Jing; Hanchen Zhang; Yanxiao Zhao; 顺天 姚; Jie Tang; Yuxiao Dong; |
| 162 | Intent-Aware Caching for Efficient LLM Serving Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our analysis of real workloads shows that user intent is a strong signal for this behavior, as different intents exhibit different continuation probabilities and inter-turn delays. Based on this insight, we propose Intent-Aware Caching (IAC), an online eviction policy that classifies requests by intent, learns lightweight reuse statistics from recent serving logs, and uses them to guide an oracle-inspired eviction decision. |
Kazi Hasan Ibn Arif; JinYi Yoon; Dimitrios S Nikolopoulos; Hans Vandierendonck; Deepu John; Bo Ji; |
| 163 | UniVR: Thinking in Visual Space for Unified Visual Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce UniVR, the first investigation into simultaneously learning complex reasoning, fine-grained physical dynamics, and long-term planning from pure visual demonstrations. |
Zhongwei Ren; Yunchao Wei; Yao Zhao; GONG WEIBO; Xiao Liu; Anran Wang; Xiangtai Li; Xiaojie Jin; |
| 164 | Brittlebench: Quantifying LLM Robustness Via Prompt Sensitivity Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce a theoretical framework for quantifying model sensitivity to prompt variants, or brittleness, that can enable us to disentangle data-induced difficulty from prompt-related variability. |
Angelika Romanou; Mark Ibrahim; Candace Ross; Kerem Oktar; Chantal Shaib; Samuel J Bell; Anaelia Ovalle; Antoine Bosselut; Jesse Dodge; Koustuv Sinha; Adina Williams; |
| 165 | City-RAG: Stepping Into A City Via Spatially-Grounded Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To this end, we present CityRAG, a video generative model that leverages large corpora of geo-registered data as context to ground generation to the physical scene, while maintaining learned priors for complex motion and appearance changes. |
Gene Chou; Charles Herrmann; Kyle Genova; Boyang Deng; Songyou Peng; Bharath Hariharan; Jason Y. Zhang; Noah Snavely; Philipp Henzler; |
| 166 | MONET: A Massive, Open, Non-redundant and Enriched Text-to-image Dataset Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce MONET, an open \emph{Apache 2.0} dataset of ${\sim}104.9$M image–text pairs collected from 2.9B raw pairs across heterogeneous open sources through successive stages of safety filtering, domain-based filtering, exact and near-duplicate removal, and re-captioning with multiple vision-language models covering short to long-form descriptions, and further augmented with synthetically generated samples. |
Benjamin Aubin; Gonzalo Iñaki Quintana; Onur Tasar; Sanjeev Sreetharan; Urszula Czerwinska; Damien Henry; Clément Chadebec; |
| 167 | Test Time Search Requires Training for Diversity Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Vector Policy Optimization (VPO), which exploits the fact that practical rewards are vector-valued, e.g., per-test-case correctness, per-criterion ratings, or per-hop credit. |
Ryan Bahlous-Boldi; Isha Puri; Idan Shenfeld; Akarsh Kumar; Mehul Damani; Sebastian Risi; Omar Khattab; Zhang-Wei Hong; Pulkit Agrawal; |
| 168 | Neural Chameleons: Language Models Can Learn to Hide Their Thoughts from Unseen Activation Monitors Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We demonstrate that fine-tuning can create **Neural Chameleons**: models capable of zero-shot evading activation monitors. |
Max McGuinness; Alex Serrano Terre; Luke Bailey; Scott Emmons; |
| 169 | Parasite Features: Causal Abstraction Through Spurious Pathways Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Suppose there are two directions in a model’s representation space where: (i) one discriminates the input based on some specific feature, and (ii) the other steers the model towards a certain output. Causal Abstraction techniques can create a pathway leveraging these unrelated directions to simulate their combined behavior in the model, which thus spuriously appears to implement a certain feature. |
Denis Sutter; Qing Yao; Sasha Boguraev; Julian Minder; Antonio Di Maio; Thomas Hofmann; Kyle Mahowald; Tiago Pimentel; |
| 170 | AstraFlow: Dataflow-Oriented Reinforcement Learning for Agentic LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This burden arises from trainer-centered control architectures and the lack of principled abstractions for RL system components. To address these limitations, we propose AstraFlow, a dataflow-oriented RL system that replaces conventional trainer-centered control with principled component abstractions. |
Haizhong Zheng; Yizhuo Di; Jiahui Wang; Shuowei Jin; Xueshen Liu; Yongji Wu; Zhuoqing Morley Mao; Ion Stoica; Jiawei Zhao; Beidi Chen; |
| 171 | Agent Explorative Policy Optimization for Agentic Multimodal Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Under standard RL recipes like GRPO, the gap manifests as two diagnostic symptoms during training: tool use is attempted on only ~30% of rollouts, and when attempted, the tool-using rollouts within a group are all-wrong on ~40% of questions, suppressing the learning signal at the tool calls that needed it. We propose AXPO (Agent eXplorative Policy Optimization): for each all-wrong tool-using subgroup, AXPO fixes the thinking prefix and resamples the tool call and its continuation, paired with uncertainty-based prefix selection. |
Minki Kang; Shizhe Diao; Ryo Hachiuma; Sung Ju Hwang; Pavlo Molchanov; Frank Wang; Byung-Kwan Lee; |
| 172 | Self-Calibrated GUI Reward Model Via Inverse Dynamic Modeling Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose I-Judge, a novel framework for Self-Calibrated Reward Modeling. |
Zeyi Sun; Shengyuan Ding; Xingpeng Xia; Jinsong Li; Xie Chen; Dahua Lin; Jiaqi Wang; |
| 173 | The Marauder’s Map: Bézier Manifolds Reveal Hidden Surfaces for Model Merging and Ensembling Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Nimbus, which learns a high-dimensional Bézier manifold connecting N specialists, going beyond the one-dimensional paths and two-dimensional sheets of prior mode connectivity work. |
Abhiram Iyer; Mark T Harnett; Sarthak Chandra; |
| 174 | DMax: Aggressive Parallel Decoding for DLLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present DMax, a new paradigm for efficient Diffusion Language Models (dLLMs). |
Zigeng Chen; Gongfan Fang; Xinyin Ma; Ruonan Yu; Xinchao Wang; |
| 175 | Reward Is Not A Universal Interface for Generative Reinforcement Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce MindRL, a reward-to-update interface controller that translates a shared reward through branch-native score objects while turning factorization, tractability, drift, and smoothness barriers into budgets over block size, serialization, anchors, clipping, reranking, and branch weights. |
Victor Shea-Jay Huang; Poodar Chu; Hongsheng Li; |
| 176 | AI Construct Lexis: An Ontology of The Hidden Assumptions in AI Evaluation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce AI Construct Lexis, a versioned ontology and interactive artifact that extracts and organizes relationships among constructs, behaviors, and measurement instruments as claimed in the literature for LLM evaluation. |
Olawale Salaudeen; Florian E. Dorner; Tom Sühr; Sang Truong; Haoran Zhang; Patrick Bissett; James Fox; Marzyeh Ghassemi; Tom Hartvigsen; Peter Hase; Augustin Kelava; Suhas Mahesh; Margaret Mitchell; Jamelle Watson-Daniels; Sanmi Koyejo; Angelina Wang; |
| 177 | Qrita: High-performance Top-k and Top-p Using Pivot-based Truncation and Selection Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose Qrita, an efficient Top-k and Top-p algorithm based on a pivot-based truncation and selection. |
Jongseok Park; Sunga Kim; Alvin Cheung; Ion Stoica; |
| 178 | DynamicVLA: A Vision-Language-Action Model for Dynamic Object Manipulation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present DynamicVLA, a latency-aware VLA model for dynamic object manipulation.To fill the missing foundation of dynamic manipulation data, we introduce the Dynamic Object Manipulation (DOM) benchmark, built with an automated collection pipeline that gathers 200K synthetic episodes across 2.8K scenes and 206 objects, and enables fast collection of 2K real-world episodes without teleoperation. |
Haozhe Xie; Beichen Wen; Jiarui Zheng; Zhaoxi Chen; Fangzhou Hong; Haiwen Diao; Ziwei Liu; |
| 179 | OneVision-Encoder: Codec-Aligned Sparsity As A Foundational Principle for Multimodal Intelligence Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Visual signals, especially videos, are highly redundant: most regions are temporally predictable, while informative changes are sparse. We hypothesize that vision encoders should allocate computation according to this non-uniform information density rather than process dense pixel grids uniformly. |
feilong tang; Xiang An; Yunyao Yan; Yin Xie; Kaicheng Yang; Yifei Shen; Yuanhan Zhang; Shikun Feng; Chunyuan Li; Changrui Chen; Huajie Tan; Ming Hu; Manyuan Zhang; Bo Li; Ziyong Feng; Ziwei Liu; Zongyuan Ge; Jiankang Deng; |
| 180 | Agents’ Last Exam Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Agents’ Last Exam (ALE), a benchmark for evaluating generalist computer-use AI agents on long-horizon professional tasks with verifiable outcomes. |
Yiyou Sun; Xinyang Han; Weichen Zhang; Yuanbo Pang; Tianyu Wang; Yuhan Cao; Yixiao Huang; Christopher Duroiu; Haoyun Zhang; Jeffrey Lin; Weishu Zhang; Tianren Zeng; Ying Yan; Bo Liu; Hanson Wen; Mingyang Xu; Xiaoyuan Liu; Zimeng Chen; Weiyan Shi; Amanda Dsouza; Vincent Chen; Yushan Li; Wenxi Deng; Huiqi Wang; Justin Xu; Tao Sun; Zhun Wang; Chris Liu; Yafei Cheng; Rongwang Hu; Aras Bacho; Shengcao Cao; Zengyi Qin; Yixiong Chen; Hengduan Fan; Hao Liu; Lin Zeng; Shashank M Bharadwaj; Litian Gong; Maojia Song; Ruheng Wang; Zongzheng Zhang; Jianhong Tu; Zhonghua Wang; Zheng Zhang; Zijiao Chen; Yanqiong Jiang; Zhendong Li; Bohan Lyu; Ma Chang; Peiran Xu; Benran Zhang; Shangding Gu; Haoyue Hua; Haoyang Li; Wanzhe Liao; Chengzhi Liu; Junbo Peng; Zechen Xu; Bo Chen; Jiayi Cheng; Yi Jiang; Maggie Kuang; Yuan Li; Youbang Pan; Ziyan Rao; Alexander Schubert; Ivan Shen; Vincent Siu; Xiatao Sun; Xiaopan Zhang; Yuchen Zhu; Ishaan S Chandok; Lei Ding; Jingxuan Fan; Andrew Glover; Jiaming Hu; Yiran HU; Wenbo Huang; Haoran Jin; Lukas Kim; Ming Liu; Alireza Rafiei; Kunyang Sun; Ting Sun; Yingze Wang; Yixin Wang; Hanwen Xing; Sihan Xu; Yuzheng Xu; Zhongxing Xu; Zhiling Yan; Boqin Yuan; Ruiqi Zhang; Yifan Zhang; Zibo Zhao; Haoyue Bai; Carlo Bosio; Joe Cavanagh; Tianxing Chen; Yipu Chen; CHENYU ZHU; Chen Dai; Stefano De Castro; Yunfu Deng; Kaustubh Dhole; Jiayuan Ding; Chenchen Du; Zhehang Du; Hao Fan; Run-Ze Fan; Hengyu Fu; Shi Gu; Yifan Gu; Baihe Huang; Baixiang Huang; Ran Jin; Xin Lan; Joseph Lee; Deren Lei; Chenyu Li; Daofeng Li; Jingyan Li; Yi Li; Yuangang Li; Zhixu Li; Yinsheng Li; Wenyu Liang; Longtai Liao; Kevin Qinghong Lin; Zeyi (Andy) Liu; Jiaming Liu; Kaiyuan Liu; Xuan Liu; Pan Lu; Yicheng Lyu; Qiuyang Mang; Kyle Montgomery; Ruoxi Ning; Jorin Overwiening; Xu Pan; Core Francisco Park; Justin Purnomo; Scott Rankin; Bixuan Ren; HaoYang Shang; Hejia Shen; JIAWEI SHEN; Shi Qiu; Tianneng Shi; Jonah So; Vladislav Susoy; Haocheng Wang; Jialu Wang; Wei Wang; Xinyu J Wang; Zehao Wang; Ruoxi Wu; Dehao Wu; Fangyu Wu; Mengyuan Wu; Yu Wu; Yuchen Wu; Yuhao Wu; Qingpo Wuwu; Weihang Xiao; Yongyi Xiong; Fan Xu; Ruiling Xu; Mingxuan Yan; Benjamin Yang; Jirong Yang; Xiaoli Yang; Yushi Yang; Haoran Ye; Xiaohu Yu; Zhengming Yu; Chenlong Zhang; Chi Zhang; Hanning Zhang; Hanwen Zhang; Junge Zhang; Kunpeng Zhang; Song Zhang; Wenjin Zhang; Wenshuo Zhang; Qijian Zhao; Yimin Zhao; Yuhaohua Zheng; Liwei Zhou; Tianyue Zhou; Sichen Zhu; Yan Zhu; Yishu Zhu; Jerry Zuo; Chonghao Cai; Helena Casademunt; Cheng Cheng; Nawen Deng; Rao Fu; Yifan Han; He Ren; Zhenyu He; Qiao Jin; Langlang Li; Yuetai Li; Lu Lu; Luqing Zhou; Subhabrata Mukherjee; Yunqi Ouyang; Yin Ren; Dawei Shi; Haoran Wu; Zhiyue Wu; Hannah Yao; Zhuoran Yi; Jenny Yu; Rhea Zhan; Hang Zhou; Shiqiao Zhu; Junfan Zhu; Zhen Dong; Ali Emami; Yanjun Gao; Haimin Hu; Erika J Schneider; Zhenglu Li; Jiachen Li; Lihui Liu; Murphy Niu; Yi Shao; Wenqi Shi; Lichao Sun; Jianxin Sun; Chenguang Wang; Xuhai Orson Xu; Huaxiu Yao; Aylin Caliskan; Jing Huang; Yang Liu; Mark W Mueller; Russell Poldrack; Sanjiv R Das; Zhiyong Lu; Radha Poovendran; Somayeh Sojoudi; Costas J Spanos; Molei Tao; Mikko Tolonen; Ting Wang; Alan Yuille; Patrick Bryant; Carl Boettiger; Arvind Rao; Tapio Schneider; Georgios N Yannakakis; Laure Zanna; Ida Sim; Tarek Zohdi; Jack Gallant; Teresa Head-Gordon; Dawn Song; |
| 181 | An Equivariance Principle for Optimizer Design: Symmetry-Compatible Updates for Embeddings, LM Heads, and MoE Routers Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We develop a symmetry-based principle for matrix-gradient optimization, showing that orthogonally equivariant optimizers, especially spectral optimizers such as stochastic spectral descent, Muon, Scion, and polar gradient methods, are better aligned with matrix layers than coordinate-wise updates. |
Tim Tsz-Kit Lau; Weijie Su; |
| 182 | NitroBox: Lightning-Fast Sandbox for Large-Scale RL Training Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose NitroBox, a novel sandbox system purpose-built for agentic RL that significantly accelerates the rollout and training process. |
Yuzhou Nie; Ruilin Zhou; Zhaorun Chen; Jingyang Zhang; Yan Shao; Hongwei Li; Bo Li; Dawn Song; Wenbo Guo; |
| 183 | TACT: Mitigating Overthinking and Overacting in Coding Agents Via Activation Steering Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we introduce **TACT** (**T**hink-**A**ct **C**alibration via activation s**T**eering), to detect and mitigate agent drift in the residual stream before it surfaces as a behavioral failure. |
Yuan Sui; Yulin Chen; Yibo Li; Xue Jiang; Yufei He; Yihong Dong; Xiaoxin He; Tianyu Gao; Bryan Hooi; |
| 184 | ParaVT: Taming The Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose **PARA-GRPO** (**P**arseability-**A**nchored and **R**atio-g**A**ted **GRPO**), which augments standard RL with two complementary mechanisms: (i) a *targeted format reward* applied only at the structural-token positions most prone to collapse, and (ii) a *per-prompt frame-budget randomization* that creates training prompts where calling the tool yields a measurable reward signal over skipping it. |
Zuhao Yang; Kaichen Zhang; Sudong Wang; Keming Wu; Zhongyu Yang; Bo Li; Xiaojuan Qi; Shijian Lu; Xingxuan Li; Lidong Bing; |
| 185 | HYDRA: Representation Harmonized Tokenization for Multimodal Generation and Understanding Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Crucially, this transition is mediated by a Generation-Semantic Bottleneck (GSB), which compresses features into a low-dimensional space to filter noise for robust synthesis, then restores dimensionality to empower complex semantic comprehension. Built upon this foundation, we present \model, a native unified framework integrating perception and generation within a single parameter space. |
Xuerui Qiu; Yutao Cui; Guozhen Zhang; Junzhe Li; Yaqi Zhao; Xiao Zhang; Yang Li; Songtao Liu; Miles Yang; Yu Shi; Zhao Zhong; Liefeng Bo; |
| 186 | Accelerating The Inference Era with AI-Driven, Globally Optimized HW/SW Co-Design Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce CHEETAH: a globally optimized, AI-driven framework for inference accelerator co-design. |
Miria Feng; Fangzhao Zhang; Adrian G Lafuente; Mert Pilanci; Azalia Mirhoseini; |
| 187 | Barycentric Guidance: Turning Foundation Image Editors Into Continuous Affective Controllers Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Meanwhile, despite producing high-quality edits, state-of-the-art foundation image editors are not continuous affective controllers: non-linear emotion representations in text encoders makes prompt interpolation poorly calibrated, so accurate and continuous VA control remains unreliable. To fill this gap, we introduce Barycentric Guidance, a training-free framework that turns pretrained foundation image editors into continuous affective controllers over VA, requiring no finetuning, extra data, or model changes. |
Harvey Mannering; Zhiwu Huang; Adam Prugel-Bennett; |
| 188 | Boosting LLM Reasoning Via Human-Inspired Reward Shaping Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: By contrast, for well-mastered problems, they focus instead on reasoning condensation and knowledge abstraction to distill concise underlying principles. Motivated by this gap, we introduce T2T(Thickening-to-Thinning), a dynamic reward framework inspired by human learning processes. |
Wenze Lin; Zhen Yang; Xitai Jiang; Xiaoteng Ma; Gao Huang; |
| 189 | From Reasoning Chains to Verifiable Subproblems: Curriculum Reinforcement Learning Enables Credit Assignment for LLM Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Correct final-answer rollouts are rare, and standard sample-level credit assignment fails to leverage partial reasoning progress embedded in unsuccessful attempts. To address this problem, we introduce **SCRL** (**S**ubproblem **C**urriculum **R**einforcement **L**earning), a curriculum reinforcement learning framework built on verifiable subproblems derived from reasoning chains. |
Xitai Jiang; Wenze Lin; Zihan Tang; Yang Yue; Shenzhi Wang; Gao Huang; |
| 190 | Beyond Sparse Captions: Aligning Slide-Level Text and Patch-Level Vision in Pathology Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose to instead leverage slide-level pathology reports which can be generated during routine diagnostic workflows in a scalable manner. |
Daniel Shao; Dongmin Bang; Luca Weishaupt; Nic G Reitsam; Sophia J. Wagner; Ming Yang Lu; Long P Le; Richard Chen; Faisal Mahmood; |
| 191 | Economy of Minds: Emerging Multi-Agent Intelligence with Economic Interactions Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: How can a population of agents self-orchestrate and self-adapt into stronger collective intelligence without centralized control? Inspired by Friedrich Hayek’s economic theory of decentralized coordination in markets, we study this question through an agent economy in which agents compete via auctions for the right to act, exchange payments, and accumulate wealth from environmental rewards. |
Zhenting Qi; Ao Qu; Huangyuan Su; Chenyu Wang; Yu Yao; Han Zheng; Kushal Chattopadhyay; Guowei Xu; Zihan Wang; Weirui Ye; Vijay Janapa Reddi; Ju Li; Paul Liang; Hima Lakkaraju; Sham Kakade; Yilun Du; |
| 192 | Designing Effective Monitor-Based Interventions for Mitigating Reward Hacking During RL Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To study how best to use monitors to mitigate reward hacking, we introduce and open source three realistic environments where Qwen3-4B reward hacks: a coding environment hackable via test overwriting, a medical chat environment hackable via sycophancy, and a biography generation environment hackable via hallucination. |
Aria Wong; Joshua Engels; Neel Nanda; |
| 193 | Model Incrimination: Investigating Whether Concerning Behavior Reflects Misalignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: A central goal of safety research is determining whether a model is misaligned. |
Gerson Kroiz; Aditya Singh; Senthooran Rajamanoharan; Neel Nanda; |
| 194 | Subliminal Learning Is Steering Vector Distillation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we show that subliminal learning is mediated by a single steering vector, i.e. a vector added to the model’s activations. |
Camila Blank; Agam Bhatia; Senthooran Rajamanoharan; Arthur Conmy; Neel Nanda; |
| 195 | Hermes: A Multi-Scale Spatial-Temporal Hypergraph Network for Stock Time Series Forecasting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: These methods face two key limitations: they do not fully consider inter-industry lead-lag interactions, and they do not model multi-scale information within and among industries. This study proposes the \textbf{Hermes} framework for stock time series forecasting that aims to improve the exploitation of industry correlation by addressing these limitations. |
Xiangfei Qiu; Liu Yang; Xiangyu Xu; Hanyin Cheng; Xingjian Wu; Rongjia Wu; Zhang Zhigang; Tu ding; Chenjuan Guo; Bin Yang; Christian S. Jensen; Jilin Hu; |
| 196 | SkillForge: Co-Evolving Skills and Agents Via Dynamic Skill Lifecycles Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose SkillForge, an agentic RL method that compiles and evolves the skill library through a fitness-driven skill lifecycle of trial, active, stable, and retired states, so that the skills and the model co-evolve throughout training. |
Yuyao Ge; Yiwei Wang; Yuchen He; Baolong Bi; Lingrui Mei; Jiayu Yao; Lizhe Chen; Shenghua Liu; |
| 197 | SkillRL: Evolving Agents Via Recursive Skill-Augmented Reinforcement Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose SkillRL, a framework that bridges the gap between raw experience and policy improvement through automatic skill discovery and recursive evolution. |
Peng Xia; Jianwen Chen; Hanyang Wang; Jiaqi Liu; Kaide Zeng; Yu Wang; Siwei Han; Yiyang Zhou; Xujiang Zhao; Haifeng Chen; Zeyu Zheng; Cihang Xie; Huaxiu Yao; |
| 198 | Paloma: Phase-Conditioned Residual Modulation for Time Series Forecasting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Motivated by this perspective and inspired by phase and amplitude modulation in classical signal processing, we propose \textit{Paloma}, a simple yet effective framework for modeling residual deviations. |
Jingru Fei; Kun Yi; Wei Fan; Qi Zhang; Zhendong Niu; |
| 199 | AI Agents May Always Fall for Prompt Injections Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Instead, through CI, we offer a principled framework for evaluating context-sensitive failures, and designing CI-aware alignment for the frontier autonomous agents. |
Sahar Abdelnabi; Eugene Bagdasarian; |
| 200 | Tio: Language Models with Parallel Streams of Thoughts, Inputs and Outputs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we show that models can be unblocked by switching from instruction-tuning for sequential message formats to instruction-tuning for **multiple, parallel streams of computation**, splitting each **role** into a separate stream. |
Guinan Su; Yanwu Yang; Xueyan Li; Jonas Geiping; |
| 201 | VideoMLA: Low-Rank Latent KV Cache for Minute-Scale Autoregressive Video Diffusion Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we present the first study of Multi-Head Latent Attention (MLA) in video diffusion. |
Hidir Yesiltepe; Jiazhen Hu; Tuna Han Salih Meral; Adil K Akan; Kaan Oktay; Hoda Eldardiry; Pinar Yanardag; |
| 202 | Rectified Policy Rollouts with Hierarchical Expert Guidance for Neural Combinatorial Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To mitigate the longstanding challenges, e.g., reward sparsity and sample inefficiency in the vast combinatorial action space which causes ineffective training, subpar performance, and poor scalability, we propose CORectifier, a novel NCO solver learned under the hierarchical gated rectification mechanism to regularize the arbitrary exploration: partial policy-predicted actions in a trajectory are probabilistically replaced with high-quality segments from reference solutions. |
Wenzheng Pan; Nuoyan Chen; Jiaxi Liu; Jiale Ma; Junchi Yan; |
| 203 | UECO: A Unified Encoder with Structure-Aware Attention Mixture Via Iterative Edge Evolving for Neural Combinatorial Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose UECO, a simple yet effective **U**nified **E**ncoder for **C**ombinatorial **O**ptimization with a structure-aware attention mixture scheme, where a lightweight module defined upon graph-based locality is interleaved with conventional (global) node-focused attention blocks to dynamically evolve edge representations and inject local topological information into attention scores. |
Wenzheng Pan; Shuyi Yan; Nuoyan Chen; Yiyang Qu; Jiale Ma; Junchi Yan; |
| 204 | SurGe: Improved Surface Geometry in Point Maps Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We evaluate our proposed method on eight common zero-shot monocular geometry estimation benchmarks and achieve state-of-the-art results, particularly improving local surface geometry. |
Karim Knaebel; Gonzalo M Garcia; Christian Schmidt; Ilya Fradlin; Lucas Nunes; Daan de Geus; Bastian Leibe; |
| 205 | XTC: Head-Aware Sampling By Excluding Top Choices Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: These strategies overlook a recurring regime of open-ended generation in which the model already assigns substantial probability to several plausible continuations yet still concentrates too much mass on the most generic choice. We introduce XTC Exclude TopChoices, a lightweight head-aware decoding operator that targets this head-ambiguity regime directly. |
Philipp E Weidmann; Allen Roush; Judah Goldfeder; Sanjay Basu; Ravid Shwartz-Ziv; |
| 206 | Dataset Distillation Via Drifting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In contrast, state-of-the-art vision approaches are increasingly building on large, pre-trained self-supervised models rather than training from scratch. In this paper, we investigate the problem of distilling datasets that enable us to optimally train \textit{linear probes} on top of such large, pre-trained vision models. |
George Cazenavette; Angelina Quan; Giannis Daras; Antonio Torralba; Vincent Sitzmann; |
| 207 | Probing for Representation Manifolds in Superposition Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This paper introduces the Manifold Probe, a supervised method for discovering representation manifolds in superposition. |
Alexander Modell; |
| 208 | What If Agents Could Imagine? Reinforcing Open-Vocabulary HOI Comprehension Through Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To address this, we propose ImagineAgent, an agentic framework that integrates cognitive mapping, tool-augmented reinforcement learning (RL), and generative world modeling for robust OV-HOI understanding.Specifically, we first propose an innovative CoT dataset named hicodet-6K for supervised fine-tuning (SFT), which effectively bridges the perception-to-cognition gap by structuring perceived entities into interaction pairs for comprehensive predictions. |
Zhenlong Yuan; Yue Wang; Jing Tang; Rui Chen; Kejin Cui; Lei Sun; Dapeng Zhang; Hongwei Yu; Chengxuan Qian; Xiangxiang Chu; Shuo Li; Yuyin Zhou; |
| 209 | Geometrically Disentangling Concept Learning from The Language Modeling Loss Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: How do language models acquire such a vast array of concepts and abilities from next-token prediction alone? We introduce an interpretability framework that exposes how this single loss relates to different concepts at different points during pretraining. |
Yupei Wang; Neil R Mallinar; Misha Belkin; Alex Warstadt; |
| 210 | Agentic Video Editing from Underspecified Requests Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present Aurora, an agentic video editing framework that pairs a tool-augmented vision-language models (VLMs) agent with a unified video diffusion transformer. |
Yongsheng Yu; Ziyun Zeng; Zhiyuan Xiao; Zhenghong Zhou; Hang Hua; Wei Xiong; Jiebo Luo; |
| 211 | PixelDiT2: Representation-Grounded Pixel Diffusion Transformers Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We argue that a key reason is the lack of an explicit representation prior: unlike latent diffusion, which denoises in a compact and structured latent space, pixel diffusion must learn denoising-friendly representations and pixel generation simultaneously from raw RGB space. To address this problem, we propose PixelDiT2, an end-to-end pixel-space diffusion model designed to decouple representation learning from pixel generation without introducing an autoencoder or latent reconstruction bottleneck. |
Yongsheng Yu; Wei Xiong; Yichen Sheng; Shiqiu Liu; Jiebo Luo; |
| 212 | WebArena-Pro: A Heterogeneous, Multimodal, Reproducible Benchmark for Web Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Moreover, existing benchmarks do not meaningfully evaluate whether agents can understand and use audio and video content embedded within web tasks. To address these gaps, we introduce WebArena-Pro, a benchmark comprising 300 tasks across 20 self-hosted web applications in six domain categories, spanning distinct interface conventions, workflows, and data models. |
Imene Kerboua; Fatemeh Pesaran Zadeh; Xing Han Lu; Weijian Qi; Alexander Miller; Junyi Song; Yunjia Tian; Dongjin Kang; Seyeon Choi; Marzia Nouri; Ewen Gueguen; Matteo Boglioni; Fengyuan Liu; Zeyi Liao; Mengqi Yuan; Yue Li; Alexandre Lacoste; Alexandre Drouin; Spandana Gella; Huan Sun; Gunhee Kim; Siva Reddy; |
| 213 | Jointly Reinforcing Diversity and Quality in Language Model Generations Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This creates a tension: while post-training improves response quality, it also sharpens output distributions and reduces the range of ideas, limiting the usefulness of LMs in creative and exploratory tasks such as brainstorming, storytelling, or problem solving. We address this challenge with Diversity-Aware Reinforcement Learning (DARLING), a framework that jointly optimizes for response quality and semantic diversity. |
Tianjian Li; Yiming Zhang; Ping Yu; Swarnadeep Saha; Daniel Khashabi; Jason Weston; Jack Lanchantin; Tianlu Wang; |
| 214 | Self-Rewarded Multimodal Coherent Reasoning Across Diverse Visual Domains Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce SR-MCR, a lightweight and label-free framework that aligns reasoning by exploiting intrinsic process signals derived directly from model outputs. |
jusheng zhang; Ningyuan Liu; Kaitong Cai; Sidi Liu; Ziliang Chen; Wenhao Wang; Keze Wang; |
| 215 | EmpathyChat: Structured Cognitive Reasoning in Empathetic Spoken Dialogue Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose EmpathyChat, a unified framework that reformulates empathetic spoken dialogue as a structured cognitive reasoning process integrating perception, mental-state reasoning, and response generation. |
Dingdong WANG; Shujie LIU; Jinyu Li; Yuxuan Hu; Yayue Deng; Yunrui Cai; Jincenzi Wu; Jianwei Yu; Helen Meng; |
| 216 | ESAM: Editing SAM3 Attention for Training-Free Referring Segmentation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Suppressing the sink alone is insufficient, as the released attention mass spreads without direction; closing the gap requires both suppressing the sink and providing patch-dependent spatial guidance that routes attention to semantically relevant content tokens. Building on these findings, we introduce eSAM, a fully training-free framework that edits SAM~3’s cross-attention without any parameter updates. |
Yaoting Wang; Yun Zhou; Hengrui Hu; Chang Liu; Henghui Ding; |
| 217 | MMDiff: Multimodal Model Diffing for Feature Discovery and Control Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While applicable to post-hoc inspection, hidden states that are decomposed into interpretable feature directions using sparse autoencoders (SAEs) do neither readily isolate which features are changed by multimodal training, nor are they directly useful for targeted control. We introduce MMDiff, a multimodal model-diffing framework that trains multimodal SAEs and turns them into feature-level interfaces for discovering and controlling multimodal behavior. |
Lachin Naghashyar; Hunar Batra; Ashkan Khakzar; Philip Torr; Ronald Clark; Christian Schroeder de Witt; Constantin Venhoff; |
| 218 | Beyond Data Scaling: Representation-Centric Pre-training for Vision-Language-Action Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose **VLAct**, a VLA-oriented VLM backbone trained with a representation-centric pre-training recipe. |
Senqiao Yang; Chengyao Wang; Yuxin Chen; Zixuan WANG; Longxiang Tang; Haokun GUI; Jinhui Ye (Hong Kong University of Science and Technology(Guangzhou))); Changsheng Lu; Xiaoyang Wu; Mingkang Zhu; Pengguang Chen; Shu Liu; Zhuotao Tian; Hengshuang Zhao; Bei Yu; Jiaya Jia; |
| 219 | Flow Map Language Models: One-step Language Modeling Via Continuous Denoising Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this setting, we show that both the flow and its associated flow map can be learned with simple cross-entropy objectives that respect the simplex geometry of the data, and we identify three distinct choices for flow map distillation whose performance we compare in practice. Using these insights, we build a flow language model (FLM), a continuous flow that matches state-of-the-art discrete diffusion baselines on the One Billion Words (LM1B) and OpenWebText (OWT) datasets. |
Chanhyuk Lee; Jaehoon Yoo; Manan Agarwal; Sheel Shah; Jerry Huang; Aditi Raghunathan; Seunghoon Hong; Nicholas Boffi; Jinwoo Kim; |
| 220 | Population-Aligned Persona Generation for LLM-based Social Simulation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose a systematic framework for synthesizing high-quality, population-aligned persona sets for LLM-driven social simulation. |
Zhengyu Hu; Jianxun Lian; Zheyuan Xiao; Max Xiong; Tianfu Wang; Teng Xiao; Yuxuan Lei; Fengqing Jiang; Kaize Ding; Ziang Xiao; Nicholas Jing Yuan; Xing Xie; Radha Poovendran; |
| 221 | Model Spec Midtraining: Improving How Alignment Training Generalizes Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce $\textbf{model spec midtraining}$ (MSM): after pre-training but before alignment fine-tuning, we train models on synthetic documents discussing their Model Spec. |
Chloe Li; Sara Price; Samuel Marks; Jonathan Kutasov; |
| 222 | UnlearningSoup: Is Repeated Tuning Necessary for Large Language Model Unlearning? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This suggests that stronger models may be recovered through an unlearning-tailored soup strategy, reducing the need for repeated tuning for further improvement or new settings. Motivated by this, we propose **UnlearningSoup**, a unified framework that provides two strategies: EfficientSoup uses binary-search-based interpolation to quickly discover a well-performing model in the early stage, where repeated tuning would otherwise make strong model selection costly. |
Puning Yang; Qizhou Wang; Junchi Yu; Bo Han; Xiuying Chen; |
| 223 | Search More, Think Less: Rethinking Long-Horizon Agentic Search for Efficiency and Generalization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose $Search\ More, Think\ Less$ (SMTL), a framework for long-horizon agentic search that targets both efficiency and generalization. |
Chengjun Yu; Shu XU; Jiaqi Wu; Qianben Chen; Tianrui Qin; Zhu; Qiexiang Wang; Jiayu Zhang; Xinpeng Liu; Xin Gui; Jingyi Cao; Yi Yao; WANG PIAOHONG; Dingfeng Shi; He Zhu; Tiannan Wang; Yuqing Wang; Maojia Song; Tianyu Zheng; Jian Yang; Jiaheng Liu; Minghao Liu; Eleanor Jiang; Wangchunshu Zhou; |
| 224 | Offline Materials Optimization with CliqueFlowmer Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we offer an alternative CMD technique based on offline model-based optimization (MBO) that fuses direct optimization of a target material property into generation. |
Jakub Grudzien Kuba; Benjamin K Miller; Sergey Levine; Pieter Abbeel; |
| 225 | Mind The Gap: Dataset and Fine-grained Evaluation for Inline Audio Descriptions Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present a comprehensive investigation into MLLM-generated inline audio descriptions.First, we introduce a professionally annotated dataset of 36 hours of videos, each averaging $\sim$5 mins.We provide guidance on the current capabilities of MLLMs for audio descriptions and release our dataset and evaluation suite to the community. |
Subhashini Venugopalan; Yingwen Tan; Taylor Roper; Jimmy Tobin; Anton Kast; Alicia Martin; Sam Sepah; Amy Pavel; |
| 226 | Evolutionary Feature Engineering for Structured Data Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Evolutionary Feature Engineering (EFE), a framework for using LLM-based evolution to discover preprocessing transformations for structured data. |
Ege Onur Taga; Yilin Zhuang; Muhammed Emrullah Ildiz; Petros Mol; Abhimanyu Das; Karthikeyan Duraisamy; Samet Oymak; |
| 227 | SimVLA: Attributing Gains in VLA Models Through Controlled Ablation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Vision-Language-Action (VLA) models often bundle architectural changes with different pretraining data, backbone scales, and optimization recipes, obscuring what actually drives progress. We introduce SimVLA, a deliberately minimal VLA—a standard vision-language backbone with a lightweight continuous-action head—as a controlled testbed for this attribution problem. |
Yuankai Luo; Woping Chen; Tong Liang; Zhenguo Li; |
| 228 | NINJA: A Navigator–Inspector Joint Architecture for Context-Efficient Issue Localization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Meanwhile, the flat exploration structure hinders the balance between file-level breadth and function-level depth. To address these limitations, we propose NINJA, a hierarchical Navigator-INspector Joint Architecture for context-efficient issue localization. |
Wenkai Fang; Shunyu Liu; Wei Luo; Zihao Wu; Yang Zhou; Tongya Zheng; Wotao Yin; Jialie Shen; Leszek Rutkowski; Mingli Song; Dacheng Tao; |
| 229 | Hearing Is Believing? Evaluating and Analyzing Audio Language Model Sycophancy with SYAUDIO Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, unlike text and vision-language sycophancy, ALM sycophancy has not been systematically studied. We therefore introduce SYAUDIO, the first benchmark dedicated to evaluating sycophancy in ALMs, consisting of 4,319 audio questions spanning Audio Perception, Audio Reasoning, Audio Math, and Audio Ethics. |
Junchi Yao; Lokranjan Lakshmikanthan; Annie Zhao; Danielle Zhao; Shu Yang; Zikang Ding; Di Wang; Lijie Hu; |
| 230 | Overcoming Catastrophic Forgetting in Visual Continual Learning with Reinforcement Fine-Tuning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We empirically trace this bottleneck to Trajectory-level Drift Agnosticism: among candidate rollouts achieving identical task rewards, the KL divergence from the preceding-task policy varies substantially, which strongly correlates with catastrophic forgetting across sequential tasks. Motivated by this insight, we propose Retention-aware Policy Optimization (RaPO), a simple yet effective RFT framework that explicitly mitigates forgetting through trajectory-level reward shaping. |
Meng Lou; Hanzhong Guo; Linwei Chen; Yizhou Yu; |
| 231 | Q-Focus: Let The Question Guide What to See in Long Videos Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Consequently, they lack the ability to dynamically preserve query-relevant details, failing to emulate the human coarse-to-fine cognitive process. To bridge this gap, we propose \textbf{Q-Focus}, a plug-and-play, two-stage visual token compression framework. |
Junbo Qiao; Xinning Chai; Dong Fang; Sifa Xie; Wenxuan Huang; |
| 232 | Auto-Rubric As Reward: From Implicit Preference to Explicit Generative Criteria Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Auto-Rubric as Reward (ARR), a framework that reframes reward modeling from implicit weight optimization to explicit, criteria-based decomposition. |
Juanxi Tian; Fengyuan Liu; Jiaming Han; Yilei Jiang; Yongliang Wu; Yesheng Liu; Haodong Li; Furong Xu; Wanhua Li; |
| 233 | PISCO: Precise Video Instance Insertion with Sparse Control Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose PISCO, a video diffusion model for precise video instance insertion with arbitrary sparse keyframe control. |
Xiangbo Gao; Renjie Li; Xinghao Chen; Yuheng Wu; Suofei Feng; Jie Yang; Qing Yin; Zhengzhong Tu; |
| 234 | Beyond [SEG] Tokens: Training-Free Video Reasoning Segmentation Via Counterfactual Inference and Contrastive Concept Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing paradigms typically compress reasoning into an isolated, opaque token (e.g., \texttt{[SEG]}) via expensive fine-tuning, or rely on training-free prompt-then-track pipelines that inevitably suffer from semantic degradation and identity switches when encountering hard distractors. To fundamentally address these vulnerabilities, we propose Counterfactual Logic and Explicit Anchoring Reasoning (\textbf{CLEAR}) framework, a training-free VRS framework inspired by human dual-system cognition. |
Tianlu Zhang; guiguang ding; Jungong Han; |
| 235 | See, Read, Compare: Candidate-Aware Verification for Agent Test-Time Scaling Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: As a result, they often miss the right candidate even when it lies within the pool. To address this gap, we propose a training-free and execution-free verifier for agent test-time scaling based on a See, Read, Compare (SRC) protocol, which treats the $K$ candidates as mutual context rather than independent inputs. |
Xinyu Ye; Yongliang Wu; Xingyu Zhu; Peng Xia; Huaijin Wu; Rui Ye; Yehui Tang; Hao Xiong; Junchi Yan; Huaxiu Yao; |
| 236 | SyncWorld: Visual Calibration Enables World Models As Zero-Shot Simulators Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce SyncWorld, an action-conditioned world model that serves as a zero-shot simulator across unseen environments without any additional training. |
Yuncong Yang; Zhengtao Han; Furkan Ozyurt; Zeyuan Yang; Han Yang; Junyi Cao; Haoyu Zhen; Yilun Du; Chuang Gan; |
| 237 | Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Gated DeltaNet-2, which generalizes both Gated DeltaNet and KDA by inheriting adaptive forgetting and channel-wise decay while addressing their shared limitation, the scalar tie between erasing and writing. |
Ali Hatamizadeh; Yejin Choi; Jan Kautz; |
| 238 | Honeyval: A Comprehensive Evaluation Framework for LLM-powered HTTP Honeypots Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: These evaluation methods are often not scalable for development, reproducible across evaluations, representative of practical attacks, and adaptable to various attacker and honeypot configurations. In this work, we bridge this gap and propose Honeyval, a comprehensive evaluation framework for LLM-powered HTTP honeypots. |
Mark Vero; Fabian Kaczmarczyck; Ivan Petrov; I Shumailov; Niels Heinen; Jamie Hayes; Tianqi Fan; Luca Invernizzi; Martin Vechev; |
| 239 | MedFlowBench: Auditing Medical Agents in Full-Study Workflows Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To study it, we introduce MedFlowBench, a full-study benchmark for VLM agents, together with MedOpenClaw, a controlled and replayable runtime in which agents operate medical imaging viewers such as 3D Slicer and QuPath. |
Weixiang Shen; Chengzhi Shen; Che Liu; Junde Wu; Jiayuan Zhu; Xiao Han; Zongyue Li; Jingpei Wu; Min Xu; Daguang Xu; Yueming Jin; Benedikt Wiestler; Daniel Rueckert; Jiazhen Pan; |
| 240 | MemPilot: Learning Transferable Latent Memory Mechanisms for LLM Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce MemPilot, a decoupled latent memory framework that dynamically guides the hidden reasoning states of frozen LLMs. |
Changlong Shi; Linhao Luo; Shigeng Chen; Guibin Zhang; Yi Chang; Shirui Pan; Chengqi Zhang; |
| 241 | HorizonComposer: Spatiotemporally Consistent Driving Video Editing with Enriched Traffic Semantics Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Instruction-guided video generative models offer a scalable solution for stress-testing autonomous driving systems by simulating diverse environmental conditions.First, we construct a hybrid dataset by combining temporally aligned real-world videos with synthetic edits, balancing foundational scene structure with unambiguous appearance signals. |
Mauricio Soroco; Yuqiu Liu; Zaid Tasneem; Francesco Pittaluga; Abhishek Aich; Wuyang Chen; Manmohan Chandraker; Ziyu Jiang; |
| 242 | Understanding Gradient Orthogonalization for Deep Learning Via Non-Euclidean Trust-Region Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we provide a theoretical analysis of this approach. |
Dmitry Kovalev; |
| 243 | Look Before You Reason: Implicit Visual Thinking for Efficient Multimodal Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While effective, their multi-turn interactions with VLMs introduce redundant token computation and slow inference, and separate region inspection from the reasoning step that depends on it. To address these issues, we propose **Lo**ok Before You **Re**ason (LoRe), an implicit visual thinking framework for efficient visual reasoning within a single-turn VLM interaction. |
Zerui Chen; Changrui Chen; Fei Ni; Jiankang Deng; |
| 244 | Rethinking Infrared Small Target Detection: A Foundation Driven Efficient Paradigm Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While large-scale visual foundation models (VFMs) exhibit strong generalization across diverse visual domains, their potential for infrared small target (SIRST) detection remains largely unexplored. To fill this gap, we systematically introduce frozen VFM representations into the SIRST task and propose a Foundation-Driven Efficient Paradigm (FDEP), a general framework compatible with diverse VFMs and SIRST networks, which improves detection accuracy without additional VFM-related inference overhead. |
Chuang Yu; Jinmiao Zhao; Yunpeng Liu; Yaokun Li; Xiujun Shu; Yuanhao Feng; Bo Wang; Xiangyu Yue; |
| 245 | Delve Into The Applicability of Advanced Optimizers for Multi-Task Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we argue that the design of mainstream optimization-based MTL methods implicitly assumes compatibility with momentum-free optimizers, which may limit the understanding of their effectiveness in modern training regimes. |
Zhipeng Zhou; Linxiao Cao; Yiming Cao; Pengcheng Wu; Peilin Zhao; Chunyan Miao; |
| 246 | HiLoc: A Hierarchical Representation Method for Spatial Localization in Multimodal Large Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We argue that both limitations stem from the formulation of spatial representation. To address this issue, we propose a simple yet novel localization representation method, HiLoc, which predicts locations in a coarse-to-fine manner with special tokens: first a grid index for coarse localization, then a cell index to refine the position within the grid. |
Evelyn Zhang; Fufu Yu; Hanjun Li; Aoqi Wu; Ke Yan; Shouhong Ding; Tianhe Ren; Xinting Hu; Xiaojuan Qi; Li Jiang; Jiaya Jia; |
| 247 | Think, Plan, Paint: Layout-Aware Reasoning for Controllable Image Generation in Unified Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Unified Multimodal Large Language Models (MLLMs) offer a promising paradigm for unifying visual understanding and generation, yet they still struggle to follow complex spatial instructions and logical constraints in controllable image generation. To address this gap, we present ATLAS, a unified framework that equips MLLMs with a human-like “Think, Plan, and Paint” paradigm. |
Junhao Liu; Jian-Wei Zhang; Tao Huang; Miles Yang; Zhao Zhong; Liefeng Bo; |
| 248 | Training Language Models Via Neural Cellular Automata Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose using neural cellular automata (NCA) to generate synthetic, non-linguistic data for *pre-pre-training* LLMs–training on synthetic-then-natural language. |
Dan Lee; Seungwook Han; Akarsh Kumar; Pulkit Agrawal; |
| 249 | PACE: Progress Actively-internalized Conditioning Execution for Long-horizon Manipulation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose **PACE**, which actively internalizes task progress as a predictive interface bridging commitment and execution, equipping the policy with a calibrated sense of the *pace* at which it advances toward task completion. |
Fei Ni; Zhuo Chen; Xianze Yao; Zerui Chen; Changrui Chen; Yifu Yuan; Shan Luo; Jianye Hao; Jiankang Deng; |
| 250 | HyperSkill: Training-Free Omnimodal GRPO Via Hypergraph-Indexed Skill-Library Evolution Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: Reinforcement learning sharpens LLM reasoning, but parameter updates are expensive, forgetful, and opaque; training-free libraries swap gradients for a textual store yet inherit … |
Haoran Luo; Shangkai Lin; Jinyang Wu; XINLIANG ZHOU; Anh Tuan Luu; |
| 251 | MMGraph-Agent: Agentic Multimodal RAG Via Cache-Inspired Multimodal Knowledge HyperGraphs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, existing methods still face challenges in high construction cost, context window limitations, and the lack of unified training across offline and online knowledge sources. To address these challenges, we propose MMGraph-Agent, a unified multimodal RAG framework based on cache-inspired multimodal knowledge hypergraphs. |
Haoran Luo; Ziyue Zhu; Shangyang Wu; Linhao Luo; Yikai Guo; Qika Lin; Fangzhi Xu; Jiapu Wang; Xiaobao Wu; Yifan Zhu; Anh Tuan Luu; |
| 252 | FluxLite: Inference-Time Proposal Control for Discrete Diffusion Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce *FluxLite*, a lightweight, training-free proposal-control framework for discrete diffusion. |
Yinuo Ren; Haoxuan Chen; Grant Rotskoff; Jiequn Han; Lexing Ying; |
| 253 | Generate in Reconstruction Space, Match in Semantic Space: Transport Geometry for One-Step Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Here, we study the benefits of SSL in generative modeling in the framework of one-step generation where the role of representation is explicit: frozen SSL features are used to match generated samples to real data. |
Hugues Van Assel; Edward De Brouwer; Saeed Saremi; Gabriele Scalia; Aviv Regev; |
| 254 | Pretraining Data Statistics Shape The Phases of Learning Entity Comparison in Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: How does data shape language model (LM) behavior throughout pretraining? We investigate this question through a case study on entity comparison, e.g., *Between France and Brazil, which country is larger? |
Yik Siu Chan; Jing Huang; Yanai Elazar; Atticus Geiger; |
| 255 | Next-Latent Prediction Transformers Learn Compact World Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce **Next-Latent Prediction** (NextLat), which extends standard next-token training with *self-supervised* predictions in the latent space. |
Jayden Teoh; Manan Tomar; Kwangjun Ahn; Edward Hu; Tim Pearce; Pratyusha Sharma; Akshay Krishnamurthy; Riashat Islam; Alex Lamb; John Langford; |
| 256 | OmniCapBench: A Deep-Structured Evaluation Framework for Fine-Grained Audio-Visual Captioning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce OmniCapBench (Omni-Video Caption Benchmark), a benchmark that reframes audio–visual caption evaluation as a deep-structured diagnostic framework. |
Zhongyu Yang; Jiale Tao; Ruitao Chen; Zuhao Yang; Yingfang Yuan; Xueliang Zhao; Auden; Kai Wang; Shuai Shao; Biao Wang; Steve Yves; Qinglin Lu; |
| 257 | Position: AI Development Should Prioritize Cognitive Security Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we argue that cognitive security must be prioritized in AI development and outline a roadmap to do so. |
Batu El; Shiye Su; Aneesh Pappu; Peggy Yin; Julie Heng; Eric Heng; Ryan Z Wang; Andreas Haupt; James Zou; |
| 258 | Dream-Cubed: Controllable Generative Modeling in Minecraft By Training on Billions of Cubes Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Dream-Cubed, a large-scale dataset of Minecraft worlds at voxel resolution, and a family of models using cubes as powerful compositional units for efficient generation of interactive 3D environments.We release the full dataset, code, and all our pretrained models, which we hope will provide a foundation for future research in efficient generative modeling for structured, interactive 3D environments. |
Tim Merino; Sam Earle; Ryunosuke Iwai; Julian Togelius; Edoardo Cetin; |
| 259 | Claudini: Autoresearch Discovers State-of-the-Art Adversarial Attack Algorithms for LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We show this pipeline to be effective in jailbreaking OpenAI’s GPT-OSS-Safeguard-20B and in prompt injections against Meta-SecAlign-70B, an adversarially robust model. |
Alexander Panfilov; Peter Romov; Igor Shilov; Yves-Alexandre de Montjoye; Jonas Geiping; Maksym Andriushchenko; |
| 260 | PACE: A Proxy for Agentic Capability Evaluation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce PACE, a framework that constructs proxy benchmarks by selecting instances from existing non-agentic evaluations whose aggregate scores most reliably predict agentic benchmark performance. |
Yueqi Song; Lintang Sutawika; Jiarui Liu; Lindia Tjuatja; Jiayi Geng; Yunze Xiao; Daniel Lee; Aditya B Soni; Vincent Lo; Xiang Yue; Graham Neubig; |
| 261 | FASTER: Rethinking Real-Time Flow VLAs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Moreover, we reveal that the standard practice of applying a constant schedule in flow-based VLAs can be inefficient and forces the system to complete all sampling steps before any movement can start, forming the bottleneck in reaction latency. To overcome this issue, we propose Fast Action Sampling for ImmediaTE Reaction (FASTER). |
Yuxiang Lu; Zhe Liu; Xianzhe Fan; Zhenya YANG; Jinghua Hou; Junyi Li; kaixin Ding; Hengshuang Zhao; |
| 262 | Faithfulness Metrics Don’t Measure Faithfulness: A Meta-Evaluation with Ground Truth Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Building on this methodology, we present **BonaFide**, a benchmark of 3,066 labeled CoTs across 13 tasks and 10 models, and use it to conduct the first systematic evaluation of prominent faithfulness metrics. |
Yoav Gur-Arieh; Ana Marasovic; Mor Geva; |
| 263 | Raven: High-Recall Sequence Modeling Via Sparse Memory Routing Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Sliding-window attention (SWA) exhibits the opposite behavior: it writes sparsely by storing explicit token representations, but only within a fixed window, so recall drops once the relevant token is evicted. Interpolating between these models, we introduce Raven, a linear-time sequence model that maintains a fixed set of memory slots and, at each step, decays and updates only a selected subset via learned, input-dependent routing. |
Arshia Afzal; Aviv Bick; Eric Xing; Volkan Cevher; Albert Gu; |
| 264 | Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Tuna-2, a native unified multimodal model that performs visual understanding and generation directly based on pixel embeddings. |
Zhiheng Liu; Weiming Ren; Xiaoke Huang; Shoufa Chen; Tianhong Li; Mengzhao Chen; Yatai Ji; Sen He; Jonas Schult; Belinda Zeng; Tao Xiang; Wenhu Chen; Ping Luo; Luke Zettlemoyer; Yuren Cong; |
| 265 | Diffusion Models Without Classifier-free Guidance Related Papers Related Patents Related Grants Related Venues Related Experts Related Code View Save Abstract: Classifier-free Guidance (CFG) is central to modern diffusion models, but operates as an inference-time correction where the training objective learns the unbiased conditional … |
Zhicong Tang; Dong Chen; Jianmin Bao; Baining Guo; |
| 266 | WorldVLN: Autoregressive World Action Model for Aerial Vision-Language Navigation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To this end, we propose WorldVLN, the first autoregressive world action model for aerial VLN. |
Baining Zhao; jiacheng xu; Weicheng Feng; Xin Zhang; Zhaolu Wang; Haoyang Wang; Shilong Ji; Ziyou Wang; Jianjie Fang; Zhiheng Zheng; Weichen Zhang; Yu Shang; Wei Wu; Chen Gao; Xinlei Chen; Yong Li; |
| 267 | MilliVid: Adaptive Latents for Long-Range Consistency in Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Transformer-based video generative models have become increasingly realistic, but long-horizon consistency remains challenging to achieve because even a few dozen frames create impractically long sequence lengths. We show that this issue can be mitigated by generating video using coarse-to-fine rollout within a multi-scale token space. |
Ishaan Chandratreya; David Charatan; Basile Van Hoorick; Sergey Zakharov; Vitor Guizilini; Phillip Isola; Vincent Sitzmann; |
| 268 | Reconstructing The Vocal Tract with Differentiable Acoustic Simulation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: The human vocal tract, the cavity consisting of one’s throat, mouth, lips opening, etc., filters one’s voice to create the sounds we know as speech. In this paper, we present a differentiable and GPU parallelizable acoustic simulator that synthesizes speech by propagating sound along an acoustic tube representation of the vocal tract, and via its gradients, solves the inverse problem: reconstructing the geometry of their vocal tract solely from the sound it produces. |
Eric Chen; Jin Woo Lee; Vincent Sitzmann; |
| 269 | Understanding Multi-View Transformers Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Multi-view transformers solve 3D vision tasks in a single feed-forward pass, yet their internal mechanisms remain poorly understood. We view their residual layers as progressively updating a latent state, and study how this state evolves during 3D reconstruction. |
Julien Gaubil; Michal Stary; Louis Martinez; Andreas Geiger; Maks Ovsjanikov; Ayush Tewari; Vincent Sitzmann; |
| 270 | Concise Reasoning Through The Lens of Lagrangian Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Most proposed approaches scalarize length and reward into a single objective, requiring coefficients or thresholds that must be re-tuned across domains and model scales. We address this brittleness by treating concise reasoning as a constrained problem that minimizes length subject to an accuracy floor, and deriving a tractable algorithm, Performance-Aware Length Update (PALU), which replaces each intractable update of Lagrangian optimization with a tractable surrogate while preserving its structural prescription. |
Chengqian Gao; Haonan Li; Taylor Killian; Jianshu She; Renxi Wang; Liqun Ma; Jorge (Zhoujun) Cheng; Shibo Hao; Zhiqiang Xu; |
| 271 | The PokeAgent Challenge: Competitive and Long-Context Learning at Scale Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present the PokéAgent Challenge, a large-scale benchmark for decision-making research built on Pokémon’s multi-agent battle system and expansive role-playing game (RPG) environment. |
Seth Karten; Jake Grigsby; Tersoo Upaa; Junik Bae; Seonghun Hong; Hyunyoung Jeong; Jaeyoon Jung; Kun Kerdthaisong; Gyungbo Kim; Hyeokgi Kim; Yujin Kim; Eunju Kwon; Dongyu Liu; Patrick Mariglia; Sangyeon Park; Benedikt Schink; Xianwei Shi; Anthony sistilli; Joseph Twin; Arian Urdu; Matin Urdu; Qiao Wang; Ling Wu; Wenli Zhang; Kunsheng Zhou; Stephanie Milani; Kiran Vodrahalli; Amy Zhang; Fei Fang; Yuke Zhu; Chi Jin; |
| 272 | MemRL: Self-Evolving Agents Via Runtime Reinforcement Learning on Episodic Memory Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, current AI agents struggle to emulate this self-evolution: fine-tuning is computationally expensive and can introduce destabilizing weight updates, increasing the risk of catastrophic forgetting, while existing memory-based methods rely on passive semantic matching that often retrieves noise. To address these challenges, we propose MemRL, a non-parametric approach that evolves via reinforcement learning on episodic memory. |
Jiaqian Wang; Shengtao Zhang; Ruiwen Zhou; Junwei Liao; Yuchen Feng; Zhuo Li; Yujie Zheng; Weinan Zhang; Ying Wen; Zhiyu li; Feiyu Xiong; Yutao Qi; Bo Tang; Muning Wen; |
| 273 | CellMSA: Context Modeling for Single-Cell Representation Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Inspired by the use of multiple sequence alignment (MSA) context in protein modeling, we propose \textbf{CellMSA}, a single-cell representation learning framework that introduces an MSA-inspired inductive bias into transcriptomic modeling. |
Suyuan Zhao; Minghao Liu; Yizhen Luo; Zaiqing Nie; |
| 274 | Instruct-Particulate: Scaling Feed-Forward 3D Object Articulation with Kinematic Control Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To address this gap, we introduce Instruct-Particulate, a model that takes a 3D mesh together with a target kinematic specification, including part descriptions, connectivity, joint types, and optional point prompts, and predicts the corresponding kinematic part segmentation and joint motion parameters.To train our model at scale, we construct a heterogeneous dataset of more than 150,000 articulated 3D objects, extending existing publicly available collections with data obtained by partially labelling other 3D models, monolithic or already decomposed into parts, with kinematic labels by means of vision-language models. |
Ruining Li; Yuxin Yao; Matt Zhou; Chuanxia Zheng; Christian Rupprecht; Joan Lasenby; Shangzhe Wu; Andrea Vedaldi; |
| 275 | Generalized Laplacian in Spectral Seriation on Manifold Data Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study high-dimensional observations $Z_i \in \mathbb{R}^p$ satisfying that $E[Z_i] = X(t_i)$, where $X(t)$ parameterizes a manifold on $0 \le t < 2\pi$, with the goal of recovering the latent phases $t_i$; a motivating example is estimating object-viewing angles from photographs. |
Ruizi Wu; Wanjie Wang; Jinchi Lv; |
| 276 | Rethinking Rubric Generation for Improving LLM Judge and Reward Modeling for Open-ended Tasks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose RRD, a practical framework for rubric refinement built on a recursive decompose–filter cycle. |
William Shen; Xinchi Qiu; Chenxi Whitehouse; Lisa Alazraki; Shashwat Goel; Francesco Barbieri; Timon Willi; Akhil Mathur; Ilias Leontiadis; |
| 277 | VisionCreator-R1: A Reflection-Enhanced Native Visual-Generation Agentic Model Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Visual content generation has moved from single-image to multi-image workflows, yet existing agents are largely plan-driven and lack a systematic reflection mechanism for correcting mid-trajectory visual errors. To address this gap, we propose VisionCreator-R1, a native visual generation agent with explicit reflection, paired with a Reflection-Plan Co-Optimization (RPCO) training methodology. |
Jinxiang Lai; Wenzhe Zhao; Zexin Lu; HUALEI Zhang; Qinyu Yang; Rongwei Quan; li zhimin; Shuai Shao; Song Guo; Qinglin Lu; |
| 278 | FireMPC: A Multi-Source Pan-Canadian Wildfire Benchmark Revealing Spatiotemporal Generalization Gaps Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: FireMPC is also the first large-scale machine-learning benchmark to embed the six knowledge-driven indicators of the Canadian Forest Fire Danger Rating System (CFFDRS), bridging knowledge-based and data-driven modeling. Building on this dataset, we propose \textbf{FWI-guided hard negative mining (FWI-HNM)} strategy. |
Zhengsen Xu; Sibo Cheng; Lanying Wang; Aryan Sharma; Linlin Xu; |
| 279 | Hierarchical World Models with Implicit Dynamics Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Implicit-HWM, a hierarchical world model that decouples state-space reachability from local control. |
Gaoyue Zhou; Yvonne Wu; Zichen Cui; Nicolas Ballas; Mahmoud Assran; Lerrel Pinto; Yann LeCun; |
| 280 | BLANP: Memory-Efficient Backpropagation-Free Local Training Via Antithetic Node Perturbation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This work introduces Backpropagation-Free Local Antithetic Node Perturbation (BLANP), a fully local training framework that eliminates gradient propagation across layer boundaries while requiring no negative samples. |
Mehrzad Karamimanesh; Marcel A. J. van Gerven; Mahyar Shahsavari; |
| 281 | Every Measurement, Every Direction, All at Once: Multimodal Flow Matching for Molecules and Spectra Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present MOSAIC, a multimodal flow matching transformer over molecular 3D structure and UV, IR, and Raman spectra. |
Thorben Prein; Elton Pan; Rafael Gomez-Bombarelli; Santiago Miret; |
| 282 | Vision-OPD: Learning to See Fine Details for Multimodal LLMs Via On-Policy Self-Distillation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We observe a regional-to-global perception gap: the same MLLM answers fine-grained questions more accurately when conditioned on evidence-centered crops than on the corresponding full images, suggesting that many failures stem from difficulty to focus on relevant evidence rather than insufficient local recognition ability. Motivated by this observation, we propose Vision-OPD (Vision On-Policy Distillation), a regional-to-global self-distillation framework that transfers the model’s own privileged regional perception to its full-image policy. |
Qianhao Yuan; Jie Lou; XingYu Li; Hongyu Lin; Le Sun; Xianpei Han; Yaojie Lu; |
| 283 | WorldMemArena: Evaluating Multimodal Agent Memory Through Action–World Interaction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing benchmarks measure recall over static dialogue, collapse memory into a single end-of-task accuracy, and reduce visual observations to captions, leaving us unable to localize failures to writing, maintenance, retrieval, or use. The rise of agent harnesses that author their own memory sharpens this gap, since we have no principled way to compare hand-designed pipelines with self-managing alternatives. |
Chengzhi Liu; Yuzhe YANG; Sophia Xiao Pu; Yepeng Liu; Lin Long; Yichen Guo; Nuo Chen; Zhaotian Weng; Yiheng Zhong; Angxiao Zong; Elena Kochkina; Simerjot Kaur; Charese Smiley; Xiaomo Liu; James Zou; Sheng Liu; Yuheng Bu; Songyou Peng; Xin Wang; |
| 284 | MARBLE: An Agent Benchmark for Spatial Reasoning and Visual Abstraction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present MARBLE, a diagnostic benchmark for spatially grounded multimodal reasoning and planning across three agent environments: Cube, Maze and Blox. |
Yulun Jiang; Naël Ouerghemi; Yekun Chai; Maria Brbic; Michael Moor; |
| 285 | Learning from Trials and Errors: Reflective Test-Time Planning for Embodied LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Embodied LLMs endow robots with high-level task reasoning, but they cannot reflect on what went wrong or why, turning deployment into a sequence of independent trials where mistakes repeat rather than accumulate into experience. Drawing upon human reflective practitioners, we introduce Reflective Test-Time Planning, which integrates two modes of reflection: \textit{reflection-in-action}, where the agent uses test-time scaling to generate and score multiple candidate actions using internal reflections before execution; and \textit{reflection-on-action}, which uses test-time training to update both its internal reflection model and its action policy based on external reflections after execution. |
Yining Hong; Huang Huang; Manling Li; Fei-Fei Li; Leonidas Guibas; Jiajun Wu; Yejin Choi; |
| 286 | FuncFormer: Circuit Representation Learning Via The Flow of Functional Propagation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we $\mbox{propose} \textbf{FuncFormer}$, a Graph Transformer that incorporates $\textbf{functional simulation}$ not as a proxy task, but as a fundamental inductive bias directly into the representation learning process. |
Yunjie Ji; Jie Wang; Zhihai Wang; Min Li; Junhua Huang; Zhihao Shi; Feng Wu; Mingxuan Yuan; Jianye Hao; |
| 287 | CoVisIT: Cross-modal Prior Guided Diffusion Model for Visible-to-Infrared Image Translation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To further address local misalignment, we introduce a Dynamic Kernel Generation Module to predict input-adaptive kernels and embed a Multi-Scale Dynamic Convolution Module into the denoising U-Net, enabling dynamic local feature aggregation for implicit alignment. |
Linfeng Tang; Tong Hu; Zizhuo Li; Hao Zhang; Han Xu; Zhenfeng Shao; Jiayi Ma; |
| 288 | ELF: Embedded Language Flows Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we show that continuous DLMs can be made effective with minimal adaptation to the discrete domain. |
Keya Hu; Linlu Qiu; Yiyang Lu; Hanhong Zhao; Tianhong Li; Yoon Kim; Jacob Andreas; Kaiming He; |
| 289 | METIS: Multi-Source Egocentric Training for Integrated Dexterous Vision-Language-Action Model Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While prior works have explored leveraging human demonstrations, they are often constrained by limited scenarios and a large visual gap between human and robots. To eliminate these limitations, we propose METIS, a vision-language-action (VLA) model for dexterous manipulation pretrained on multi-source egocentric datasets. |
Yankai Fu; Ning Chen; Junkai Zhao; Shaozhe Shan; Guocai Yao; Pengwei Wang; Zhongyuan Wang; Shanghang Zhang; |
| 290 | AutoScientists: Self-Organizing Agent Teams for Long-Running Scientific Experimentation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce AutoScientists, a decentralized team of AI agents for long-running computational scientific experimentation. |
Shanghua Gao; Ada Fang; Marinka Zitnik; |
| 291 | TerraMesh-Masks: Open‑Vocabulary Segmentation for Earth Observation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We, therefore, introduce TerraMesh-Masks, a large-scale open-vocabulary segmentation dataset for EO, comprising 16 million binary masks aligned with nine million multimodal TerraMesh samples and annotated using OpenStreetMap and Overture-derived semantics.In addition, we provide an expert-reviewed evaluation set of 523 samples covering over 200 classes to enable robust evaluation.We release the dataset publicly under a permissive license at https://huggingface.co/datasets/ibm-esa-geospatial/TerraMesh-Masks. |
Benedikt Blumenstiel; Hugues Devimeux; Johannes Jakubik; Konrad Schindler; |
| 292 | MAMQ-Net: A Robust Framework Leveraging Multi-level Tamper-aware Queries and Complementary Representations of Tampering Features for Progressive Image Forgery Localization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Inspired by the way humans continuously integrate multi-faceted knowledge to understand the world, we propose MAMQ-Net, which contains a novel Multi-stage Alternating Feature Extraction and Interaction architecture. |
Minghao Jia; Zhuoyi Zhang; Jiwei Zhang; Siwei Wang; Feifei Kou; Lei Shi; ShaoZhang Niu; Jiaming Pei; |
| 293 | One Unified Representation: Resolving The Appearance-Semantics Dilemma Via Structural Regularization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This failure arises because the reconstruction loss imposes an appearance prior that erodes semantic structure. To resolve this conflict, we propose Semantic Structure Regularization (SSR). |
RUIQI YANG; Fei Zhou; Yi Zhang; Liang Li; Xiangyu Yue; |
| 294 | Majority-of-Three Is An Optimal PAC Learner Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Their main open question was whether this algorithm is optimal in the classical *high-probability* regime. In this work, building on their ideas, we answer this question affirmatively. |
Grigoris Velegkas; |
| 295 | RoboWits: Unexpected Challenges for Robotic Creative Problem Solving Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To enable scalable construction of high-quality reasoning-centric unexpected scenarios, we propose an automated task generation pipeline formulated as a multi-agent cooperative framework, comprising agents for seed task generation and verification, metric generation, scene generation, and task mutation. |
Chunru Lin; Hongxin Zhang; Fenghao Yu; Zhehuan Chen; Tom Griffiths; Yejin Choi; David Held; Chuang Gan; |
| 296 | VeruSAGE-Bench: A Benchmark Suite for Rust System Verification Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Large language models (LLMs) have shown impressive capability to understand and develop code. |
Chenyuan Yang; Natalie Neamtu; Chris Hawblitzel; Jacob R Lorch; Shan Lu; |
| 297 | Active Learning of Conditional Generative Models Via The Transport Neural Tangent Kernel Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We first show that the Wasserstein test risk of a conditional generative task decomposes into a mean term and a shape term, demonstrating why mean-based active learning is blind to distributional shape. To capture both, we propose the transport Neural Tangent Kernel (tNTK), a tractable kernel that measures how sensitive a generative model’s predicted distribution is with respect to its parameters. |
Jayoung Ryu; Kyunghyun Cho; Romain Lopez; |
| 298 | Autoregressive Appearance Prediction for 3D Gaussian Avatars Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a 3D Gaussian Splatting avatar model with a spatial MLP backbone that is conditioned on both pose and an appearance latent. |
Michael Steiner; Zhang Chen; Alexander Richard; Vasu Agrawal; Markus Steinberger; Michael Zollhöfer; |
| 299 | Geometric Inductive Biases for Semi-Supervised Equalization: The Constellation-Aware Transformer Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce the *Constellation-Aware Transformer* (CAT), a novel architecture that explicitly injects geometric inductive biases into the equalization process. |
Avi Caciularu; |
| 300 | Preserving Exploration for LLM Reasoning Via Mean Order-Statistic Alignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: A natural remedy is to impose token-wise KL divergence between the fine-tuned and base models; however, we argue that such objective introduces a fundamental mismatch, where token-wise KL operates locally at individual token positions, while exploration is a global behavior that emerges over sequences as a whole. Motivated by this insight, we propose Mean Order-Statistic Alignment (MOSA), which preserves exploration by aligning the fine-tuned model’s global behavior to that of the base model. |
Ruotian Peng; Yi Ren; Long-Fei Li; Yandong Wen; |
| 301 | Advancing Affordance-Grounded Creative Tool Use in Large Multimodal Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our evaluation shows that current LMMs often fail to sustain grounded exploration: they overlook relevant entities, under-examine critical parts, or hallucinate unsupported attributes. To address these failures, we propose **affordance-grounded alignment**, framing creative tool use as a preference learning problem. |
Cheng Qian; Hyeonjeong Ha; Jiayu Liu; Jeonghwan Kim; Emre Can Acikgoz; Bingxuan Li; Kunlun Zhu; Jiateng Liu; Aditi Tiwari; Zhenhailong Wang; Xiusi Chen; Heng Ji; |
| 302 | Agent MechSuits : Mechanistic Subspace Safety Steering for Multi-Turn CLI Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we investigate the safety of multi-turn CLI agents from an internal perspective. |
Weidi Luo; Qiming Zhang; Mingyu Jin; Jie Cai; Yihao Quan; Chaowei Xiao; Jingcheng Niu; Zhen Xiang; |
| 303 | Nemotron-Cascade: Scaling Cascaded Reinforcement Learning for General-Purpose Reasoning Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose cascaded domain-wise reinforcement learning (Cascade RL) to develop Nemotron-Cascade, capable of operating in both \emph{instruct} and deep \emph{thinking} modes, without any performance gap relative to a thinking-only counterpart. |
Boxin Wang; Chankyu Lee; Nayeon Lee; Sheng-Chieh Lin; Wenliang Dai; Yang Chen; Yangyi Chen; Zhuolin Yang; Zihan Liu; Mohammad Shoeybi; Bryan Catanzaro; Wei Ping; |
| 304 | VeriGraph: Towards Verifiable Data-Analytic Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In particular, deterministic computations over raw data and semantic deductions over natural-language claims are often entangled in an unstructured stream, leaving numerical conclusions hard to reproduce and qualitative judgments hard to inspect. To address this, we propose VeriGraph, a traceable neuro-symbolic reasoning framework that enables agents to construct an explicit heterogeneous evidence directed acyclic graph (DAG) during execution. |
Jiajie Jin; Zhao Yang; WenLe Liao; Yuyang Hu; Guanting Dong; Xiaoxi Li; Yutao Zhu; Zhicheng Dou; |
| 305 | Hierarchical Denoising For Multi-Step Visual Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Bridging this gap, we propose HDR (Hierarchical Denoising for Visual Reasoning), a unified framework for multi-step reasoning by integrating hierarchical latents into the causal video generation process.We construct a level-stratified multi-step video reasoning benchmark with out-of-distribution cases, covering six tasks: maze navigation, Tower of Hanoi, one-line drawing, sliding puzzle, Sokoban, and water pouring. |
Zezhong Qian; Xiaowei Chi; Chak-Wing Mak; Tianze Zhou; Ruibin Yuan; Yuhan Rui; Hengzhe Sun; Zhuoqun Wu; Yuming Li; Siyuan Qian; Sirui Han; Shanghang Zhang; |
| 306 | AgenTracer-v2: Agentic Failure Tracer for LLM Agentic Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, most existing approaches remain constrained to passive, full-trajectory analysis, which limits their effectiveness in ultra-long and dynamically evolving trajectory attributions. To address this limitation, we introduce AgenTracer-v2, an autonomous, multi-hop failure attribution agent that reconceptualizes tracing as an active, tool-augmented diagnostic process. |
Guibin Zhang; Haoyu Lu; Junhao Wang; He Zhu; Kun Wang; Wangchunshu Zhou; Shuicheng Yan; |
| 307 | LLM-enabled Applications Require Systematic Threat Monitoring Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: LLM-enabled applications are rapidly reshaping the software ecosystem by using large language models as core reasoning components for complex task execution. |
Yedi Zhang; Haoyu Wang; XIANGLIN YANG; Jin Song Dong; Jun Sun; |
| 308 | Strategic Decision Support for AI Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Departing from the classical view of decision support, we revisit its two basic principles, the cost–value tradeoff of seeking support and the role of uncertainty quantification, in a setting where AI agents are the central actors. We propose a framework for **strategic decision support** for AI agents through an optimization problem that minimizes support usage subject to controlling a counterfactual missed-support error: the probability that the agent acts alone on instances where support would have materially improved its output. |
Shayan Kiyani; Sima Noorani; George J. Pappas; Hamed Hassani; |
| 309 | Magnitude-preserving Layers Enable Efficient GANs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In sum, our work presents a new SOTA in FID/parameter efficiency derived from simple adversarial training with well-behaved magnitude preserving layers. |
Nick Huang; Jackson Woodleigh; Aaron Gokaslan; Xinjie Yi; James Tompkin; |
| 310 | MiroEval: Benchmarking Multimodal Deep Research Agents in Process and Outcome Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Most also offer limited multimodal coverage, rely on synthetic tasks that do not reflect real-world query complexity, and cannot be refreshed as knowledge evolves. To address these gaps, we introduce **MiroEval**, an evaluation framework for deep research systems, accompanied by a 100-task benchmark (70 text-only, 30 multimodal) constructed via a dual-path pipeline that supports periodic updates, enabling a live and evolving setting. |
Fangda Ye; Yuxin Hu; Pengxiang Zhu; Yibo Li; Ziqi Jin; Yao Xiao; Yibo Wang; Lei Wang; Zhen Zhang; Lu Wang; Yue Deng; Bin Wang; Yifan Zhang; Liangcai Su; Xinyu Wang; He ZHAO; CHEN WEI; Qiang Ren; Bryan Hooi; Bo An; Shuicheng Yan; Lidong Bing; |
| 311 | Benchmarking Graph Self-Supervised Learning for Node-Level Tasks: Insights and Strong Baseline Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To verify the effectiveness of this approach, we propose a simple generative method, GrASP, which reconstructs both graph structure and node features. |
Dmitry Eremeev; Gleb Bazhenov; Oleg Platonov; Artem Babenko; Liudmila Prokhorenkova; |
| 312 | DocAtlas: Long-Document Understanding As Mutable-State Interaction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present DocAtlas, a system that treats long-document understanding as a mutable-state information-seeking process. |
Hongchen Wei; Yuanzhe Wang; Bei Liu; Yifan Yang; Qi Dai; Kai Qiu; Yunsheng Li; Dongdong Chen; Chong Luo; Zhenzhong Chen; Baining Guo; |
| 313 | XL-DocBench: Benchmarking Evidence-Grounded Extra-Long Document Understanding Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce \textbf{XL-DocBench}, a fully human-verified benchmark for extra-long document understanding, with 1519 retained questions from six professional domains and contexts up to 2,303 pages.We build the benchmark with a tree-guided synthesis pipeline followed by artifact filters and full verification by 194 human experts. |
Hongchen Wei; Yuanzhe Wang; Bei Liu; Yifan Yang; Qi Dai; Ruichun Ma; Kai Qiu; Yunsheng Li; Dongdong Chen; Chong Luo; Zhenzhong Chen; Baining Guo; |
| 314 | The Piggyback Hypothesis of Generalization: Explaining and Mitigating Emergent Misalignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose the Piggyback Hypothesis: the chat-template prefix, shared across all user queries, can piggyback the finetuned behavior onto out-of-domain queries. |
Jiachen Zhao; Zhengxuan Wu; Aryaman Arora; Yiyou Sun; David Bau; Weiyan Shi; |
| 315 | GenEnv: Difficulty-Aligned Co-Evolution Between LLM Agents and Environment Simulators Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Training capable Large Language Model (LLM) agents is critically bottlenecked by the high cost and static nature of real-world interaction data. We address this by introducing GenEnv, a framework that establishes a difficulty-aligned co-evolutionary game between an agent and a scalable, generative environment simulator. |
Jiacheng Guo; Ling Yang; Peter Chen; Qixin Xiao; Yinjie Wang; Xinzhe Juan; Ke Shen; Mengdi Wang; |
| 316 | Gen-Searcher: Reinforcing Agentic Search for Image Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we present Gen-Searcher, as the first attempt to train a search-augmented image generation agent, which performs multi-hop reasoning and search to collect the textual knowledge and reference images needed for grounded generation. |
Kaituo Feng; Manyuan Zhang; Shuang Chen; Yunlong Lin; Kaixuan Fan; Yilei Jiang; Hongyu Li; Dian Zheng; Chenyang Wang; Xiangyu Yue; |
| 317 | Reasoning-Aware Relational Representation Learning for Open-Vocabulary Scene Graph Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Open-Vocabulary Scene Graph Generation (OV-SGG) requires models to recognize visual relationships beyond the training vocabulary, yet existing methods often rely on dataset-specific object–predicate co-occurrence patterns. Under long-tailed distributions and noisy object localization, such reliance leads to biased, entangled relation representations, limiting both rare-relation recognition and generalization to unseen predicates.To address these challenges, we propose RaRe, a reasoning-aware relational representation learning framework that transforms MLLM-generated chain-of-thought (CoT) rationales into structured relation embeddings. |
Jiawei Xiao; Boya Wang; Longtian Qiu; Shan Ning; Xuming He; |
| 318 | ThinkSafe: Self-Generated Safety Alignment for Reasoning Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Guided by this analysis, we propose ThinkSafe, a self-generated alignment framework that restores safety without external teachers. |
Seanie Lee; Sangwoo Park; Yumin Choi; Gyeongman Kim; Minki Kang; Jihun Yun; Dongmin Park; Jongho Park; Sung Ju Hwang; |
| 319 | Intelligence Per Watt: Measuring Intelligence Efficiency of Local AI Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose intelligence per watt (IPW), task accuracy divided by unit of power, as a unified metric for assessing both the capability and efficiency of local inference across model-accelerator configurations. |
Jon Saad-Falcon; Avanika Narayan; Hakki Akengin; J. W Griffin; Herumb Shandilya; Adrian G Lafuente; Medhya Goel; Rebecca Joseph; Shlok Natarajan; Etash Guha; Shang Zhu; Ben Athiwaratkun; Azalia Mirhoseini; Christopher Ré; |
| 320 | Escaping The Cognitive Well: Efficient Competition Math with Off-the-Shelf Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we present an inference pipeline that attains best-in-class performance on IMO-style math problems at an average inference cost orders of magnitude below competing methods while using only general-purpose off-the-shelf models. |
Xingyu Dang; Rohit Agarwal; Rodrigo Porto; Anirudh Goyal; Liam Fowl; Sanjeev Arora; |
| 321 | LiSA: Lifelong Safety Adaptation Via Conservative Policy Induction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This creates a practical gap: guardrails must adapt to their own operating environments, yet deployment feedback is typically limited to sparse, noisy user-reported failures, and repeated fine-tuning is often impractical. To address this gap, we propose **LiSA** (**Li**felong **S**afety **A**daptation), a conservative policy induction framework that improves a fixed base guardrail through structured memory. |
Minbeom Kim; Lesly Miculicich; Bhavana Dalvi Mishra; Mihir Parmar; Phillip Wallis; Bharath Chandrasekhar; Kyomin Jung; Tomas Pfister; Long T. Le; |
| 322 | Not All Proofs Are Equal: Evaluating LLM Proof Quality Beyond Correctness Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While this proof quality is subjective and depends on the reader and context, many of its components are concrete and broadly valued. In this work, we identify such components and introduce ProofRank, a benchmark curated from challenging mathematical competitions. |
Ivo Petrov; Jasper Dekoninck; Dimitar I. Dimitrov; Martin Vechev; |
| 323 | Step-dLLM: Adaptive Step-aware Sparse Attention for Efficient Diffusion LLM Inference Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We argue both choices ignore key structural properties of attention in dLLMs: empirically, attention patterns are only locally consistent across steps and shift abruptly at certain transitions, and different heads concentrate their attention mass to very different extents. |
Zhichen Zeng; Xichong zhang; Junpan Wu; Yifei Zuo; Chi-Chih Chang; Jiayi Wang; Maohua Nie; Ji Liu; Ang Li; Banghua Zhu; |
| 324 | Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce ${\bf Flash-dLLM}$, a training-free inference acceleration framework for fast and memory-efficient dLLMs. |
Quan Nguyen-Tri; Mukul Ranjan; Zhiqiang Shen; |
| 325 | Difference of Convex Programming in The Wasserstein Space with Applications to MMD Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: A widely used approach is to perform the optimization directly over the Wasserstein space, but many objective functionals of practical interest are non-convex along Wasserstein geodesics, making the analysis of standard first-order methods challenging. In this work, we study a class of objectives over the Wasserstein space that admit a difference-of-convex (DC) decomposition and we lift the classical convex-concave procedure (CCCP) to this setting. |
Clément Bonet; Pierre-Cyril Aubin-Frankowski; Youssef Mroueh; |
| 326 | Benchmarking Optimizers for Large Language Model Pretraining Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This study presents a comprehensive evaluation of recent optimization techniques across standardized LLM pretraining scenarios, systematically varying model size, batch size, and training duration. |
Andrei Semenov; Matteo Pagliardini; Martin Jaggi; |
| 327 | Dynamic Context Modeling for Longitudinal Mental Health Monitoring Under Distribution Shift Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: As a result, approaches that address these distribution shifts by updating model parameters are difficult to apply. To address this gap, we propose DyCon (Dynamic Context Modeling), a dual-loop framework that personalizes by building and evolving a separate natural-language personal context for each user, without updating any model parameters. |
Seungwan Jin; Taehyung Noh; Junghyun Kim; Uichin Lee; Kyungsik Han; |
| 328 | Learning Agentic Policy from Action Guidance Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose \textsc{ActGuide-RL}, which injects action data as plan-style reference guidance, enabling the agentic policy to overcome reachability barriers to reward states. |
Yuxiang Ji; Zengbin Wang; Yong Wang; Yang; Ziyu Ma; Guanhua Chen; Zonghua Sun; Liaoni Wu; Xiangxiang Chu; |
| 329 | Procedural Memory Distillation: Online Reflection for Self-Improving Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Across episodes and epochs, the model repeatedly encounters related problems under a changing policy, producing cross-episode signals that episode-local updates cannot capture: which strategies consistently pass verification, which failure modes persist, which patterns recur. We propose Procedural Memory Distillation (PMD), which converts these cross-episode signals into reusable procedural memory and distills it into the policy’s weights during training. |
Ye Liu; Srijan Bansal; Bo Pang; Yang Li; Zeyu L Liu; Yifei Ming; Zixuan Ke; Shafiq Joty; Semih Yavuz; |
| 330 | Can LLM Agents Respond to Disasters? Benchmarking Heterogeneous Geospatial Reasoning in Emergency Operations Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we introduce **D**isaster **O**perational **R**esponse **A**gent benchmark (**DORA**), the first agentic benchmark for end-to-end disaster response: 515 expert-authored tasks across 45 real-world disaster events spanning 10 types, paired with expert-verified, replayable gold trajectories totaling 3,500 tool-call steps. |
Junjue Wang; Weihao Xuan; Heli Qi; Pengyu Dai; Kunyi Liu; Hongruixuan Chen; Zhuo Zheng; Junshi Xia; Stefano Ermon; Naoto YOKOYA; |
| 331 | Sampling Is Not Curiosity: Why LLM Agents Should Investigate Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Thus, we propose a framework defining curiosity in LLM agents, reserving the term for systems that satisfy it. |
Alfonso Amayuelas; Piotr Piękos; Xin Wang; William Yang Wang; Jürgen Schmidhuber; Alex Pentland; |
| 332 | Noisy Data Is Destructive to Reinforcement Learning with Verifiable Rewards Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Recent studies suggest that improved RLVR algorithms allow models to learn effectively from incorrect annotations, achieving performance comparable to learning from clean data. In this work, we show that these findings are invalid because the claimed 100% noisy training data is contaminated with clean data. |
Yuxuan Zhu; Daniel Kang; |
| 333 | Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we present Colosseum, a framework for auditing LLM agents’ collusive behavior in multi-agent settings. |
Mason Nakamura; Abhinav Kumar; Saswat Das; Sahar Abdelnabi; Saaduddin Mahmud; Ferdinando Fioretto; Shlomo Zilberstein; Eugene Bagdasarian; |
| 334 | PIVOT: A Unified Agentic Framework for Streaming Long-Video Understanding Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce PIVOT (Planning over Incremental Video Observations for Timely Answering), a unified agentic framework that incrementally updates bounded memory from the observed stream and uses a planner-controlled action loop to retrieve relevant memory, analyze current evidence, reflect on candidate answers, and decide whether to answer or defer. |
Qiushi Lyu; Qianlan Yang; Ziqi Pang; Yu-Xiong Wang; Liangyan Gui; |
| 335 | VerifyMAS: Hypothesis Verification for Failure Attribution in LLM Multi-Agent Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Moreover, directly predicting failures induces a large combinatorial search space, hindering fine-grained attribution. To address these challenges, we propose VerifyMAS, a hypothesis verification framework for agent failure attribution. |
Hezhe Qiao; Hanghang Tong; Ee-peng Lim; Bing Liu; Guansong Pang; |
| 336 | Generalization Analysis of Biased Stochastic Gradient Methods for Minimax Problems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we present the first framework to study the stability and generalization of biased stochastic gradient methods for minimax problems. |
Shuang Zeng; Yunwen Lei; Yiming Ying; |
| 337 | Thinking with Images As Continuous Policy: Numerical Visual Chain-of-Thought Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose Numerical Visual Chain-of-Thought (NV-CoT), a framework that enables MLLMs to reason over images using continuous numerical coordinates. |
Kesen Zhao; Beier Zhu; Junbao Zhou; Xingyu Zhu; Zhongqi Yue; Hanwang Zhang; |
| 338 | Accelerated Image Editing Via Consistency-Aware Source Token Pruning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While effective, this design incurs substantial computational overhead in attention layers. To mitigate this inefficiency, we propose {\em TokenDrop}, an efficient editing framework that shift the computational burden from dense source-token conditioning to a lightweight regularized sampling update. |
Jeongsol Kim; Jong Chul Ye; Yufei Wang; Jian Wang; |
| 339 | On Reparameterizing The Score Function Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce History-Reparameterized (HR) policy gradients, which replace the frozen prefix in the score function with a differentiable reparameterized or relaxed history, allowing gradients of downstream log-probabilities to propagate backward through the generated trajectory. |
Mohammad Hossein Amani; Leonardo Bocchieri; Maxime Peyrard; Clement Hongler; Robert West; |
| 340 | GISA: A Benchmark for General Information-Seeking Assistant Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Moreover, these benchmarks tend to focus on either locating specific information or aggregating information from multiple sources, while relying on static answer sets prone to data contamination. To bridge these gaps, we introduce GISA, a benchmark for General Information-Seeking Assistants comprising 373 human-crafted queries reflecting authentic information-seeking scenarios. |
Yutao Zhu; Xingshuo Zhang; Maosen Zhang; Jiajie Jin; Liancheng Zhang; Xiaoshuai Song; Kangzhi Zhao; Wencong Zeng; Ruiming Tang; Han Li; Ji-Rong Wen; Zhicheng Dou; |
| 341 | Bridging Textual Profiles and Latent User Embeddings for Personalization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing methods typically employ either supervised latent user embeddings, which are effective for retrieval but difficult to interpret, or textual user profiles, which are interpretable but challenging to optimize for downstream utility due to lack of direct supervision. To bridge this gap, we present BLUE, a reinforcement learning framework that unifies these two forms of user representation by aligning language-based user profiles with embedding-based recommendation objectives. |
Zhaoxuan Tan; Xiang Zhai; Yan Zhu; Meng Jiang; Mohamed Hammad; |
| 342 | Elicited Adaptation: Auditable Localized Fairness Via Pairwise Queries Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce \emph{elicited localized fairness}: an end-to-end pipeline that elicits per-community Mahalanobis metric–tolerance specifications $\cF_k = (d_k, \eps_k)$ from stakeholders via two phases of pairwise queries (similarity, then tolerance acceptability), trains a localized classifier (\textsc{LocAdapt}), and audits the deployed model on a held-out tolerance fold. |
Shrey Shah; |
| 343 | Eliciting Secret Knowledge from Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study secret elicitation: discovering knowledge that an AI possesses but does not explicitly verbalize. As a testbed, we train three families of large language models (LLMs) to possess specific knowledge that they apply downstream but deny knowing when asked directly. |
Bartosz Cywiński; Emil Ryd; Rowan Wang; Senthooran Rajamanoharan; Neel Nanda; Arthur Conmy; Samuel Marks; |
| 344 | KG-Guard: Graph-Based Hallucination Detection for Knowledge Base Question Answering Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We formulate hallucination detection in KBQA as an answer-node classification problem and propose a lightweight graph-based framework that treats the answering LLM as a black box. |
Albert Sawczyn; Piotr Bielak; Tomasz Kajdanowicz; |
| 345 | Reading Between The Dots: Decoding Hidden Computation Across Filler Tokens Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: Frontier LLMs can perform multi-step reasoning over content-free filler tokens like dots or counting sequences, producing correct answers with no visible chain-of-thought (CoT). … |
Kaley Brauer; Claudio Mayrink Verdun; Samuel Marks; |
| 346 | Stop Quantum Machine Learning; Do AI-for-Quantum Instead Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This position paper argues that the NeurIPS community should prioritize AI-for-Quantum (AI4Q) over Quantum Machine Learning (QML) on classical data, and should review the two directions by different evidentiary standards. |
ziqing Guo; Ziwen Pan; |
| 347 | UniClawBench: A Universal Benchmark for Proactive Agents on Real-World Tasks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Moreover, their scenario-based task taxonomies mix multiple model capabilities within the same task category, making it difficult to identify the root causes of agent failures. To address these limitations, we introduce UniClawBench, the first capability-driven benchmark designed to evaluate proactive agents in dynamic, real-world settings. |
Zhekai Chen; CHENGQI DUAN; Kaiyue Sun; Bohao Li; Yuqing Wang; Manyuan Zhang; Xihui Liu; |
| 348 | SAMPPO: Structure-Aware Mirror Proximal Policy Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In “semantically dense” regions, where candidate tokens act as semantic synonyms within the current context, noisy advantage estimates can destabilize the fixed trust region, leading to *policy collapse*, a severe loss of generative diversity and policy entropy. To resolve this, we propose *Structure-Aware Mirror Proximal Policy Optimization (SAMPPO)*, which introduces an adaptive trust region governed by local semantic density. |
Corinna Cortes; Mehryar Mohri; Yutao Zhong; |
| 349 | Rethinking Language Model Scaling Under Transferable Hypersphere Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce HyperP (Hypersphere Parameterization), the first framework for transferring optimal learning rates across model width, depth, training tokens, and Mixture-of-Experts (MoE) granularity under the Frobenius-sphere constraint with the Muon optimizer. |
Liliang Ren; Yang Liu; yelong shen; Weizhu Chen; |
| 350 | MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To this end, we present \textbf{MemSkill}, which reframes these operations as learnable and evolvable memory skills, structured and reusable routines for extracting, consolidating, and pruning information from interaction traces. |
Haozhen Zhang; Quanyu Long; Jianzhu Bao; Tao Feng; Weizhi Zhang; Haodong Yue; Wenya Wang; |
| 351 | DeltaPrompts: Escaping The Zero-Delta Trap in Multimodal Distillation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We quantify this gap through $\textit{answer divergence}$ ($\Delta$), demonstrating that non-zero divergence is critical for effective scaling. Building on this insight, we propose a staged synthesis pipeline that repurposes existing datasets as seeds, actively targeting student failure modes to produce better prompts. |
Jaehun Jung; Hyunwoo Kim; Brandon Cui; Ximing Lu; David Acuna; Prithviraj Ammanabrolu; Yejin Choi; |
| 352 | Hide-and-Seek in Trajectories: Discovering Failure Signals for VLA Runtime Monitoring Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose \textbf{Hide-and-Seek}, a framework that formulates VLA failure detection as a coarsely supervised learning problem. |
Seongheon Park; Wendi Li; Changdae Oh; Samuel (Min-Hsuan) Yeh; Zsolt Kira; Michael Hagenow; Sharon Li; |
| 353 | Voice Cloning Is Actually Style Transfer Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Instead, we find that widely-used voice cloning models systematically apply style transfer to source voices. |
Kaitlyn Zhou; Federico Bianchi; Martijn Bartelds; Anna Pot; Yongchan Kwon; James Zou; |
| 354 | AI Agents Push Humans Out of The Loop Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To address this, a top priority in the advancement of AI agents should be supporting the situated goals and cognitive requirements of effective human oversight, treating the human needs of overseers at the same level of importance as AI agent capability. To put this idea into practice, we connect work on automation and human-computer interaction to AI agent processes, outlining design-level affordances and organizational protocols that (1) support overseers in exercising critical judgement and (2) counteract the skill atrophy that arises from extended use of automation. |
Margaret Mitchell; Avijit Ghosh; Samir Passi; |
| 355 | StableVQ: Practical Guidelines for Stable Vector-Quantized Tokenizer Training Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We argue that the root cause lies in the entanglement of the Encoder–Decoder and Codebook training: because neither module can reliably fulfill its own responsibility in isolation, the system can only function when the two subsystems happen to cooperate—a fragile condition that breaks down precisely when training is most stressed. We propose StableVQ, which revisits the proper learning objective of each module and resolves the problems that arise when each is trained to fulfill its own role independently. |
Bao Tang; Jiahao Guo; Haoxiang Cao; Wenyu Liu; Changqian Yu; Kun Gai; Xinggang Wang; |
| 356 | URDF-Anything+: End-to-End Generation for Simulation-Ready Articulated Assets Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present URDF-Anything+, an end-to-end autoregressive diffusion framework that generates simulation-ready URDF models directly from a single RGB image. |
Zhuangzhe Wu; Yue Xin; Chengkai Hou; Minghao Chen; Yaoxu Lyu; Jieyu Zhang; Shanghang Zhang; |
| 357 | On The Importance of Gating: Memorization Vs. In-Context Learning in State Space Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While SSMs excel in certain settings like length generalization, they continue to lag behind Transformers on tasks that require in-context learning and precise retrieval, slowing their adoption for large-scale language modeling. In this work, we demonstrate that both the success and failure of SSMs in these domains can be explained by studying the role of the gating mechanism, a prevalent component in modern recurrent networks. |
William Tong; Aryo Lotfi; Emmanuel Abbe; Konstantinos Vaggelakos; Vishnu Banna; Etai Littwin; Joshua Susskind; Eran Malach; |
| 358 | Large Language Model Failures from Hallucination to Homogenization Are Different Facets of Miscalibration Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This paper argues that calibration should become a primary evaluation and optimization target for Large Language Models (LLMs), because diverse failures, from confident hallucinations to collapsed diversity to brittle safety refusals, are best understood as different facets of a single underlying problem: miscalibration. We propose a unified framework organized around four types of miscalibration: probabilistic (the model cannot match requested probability distributions), semantic (confidence is misaligned with factual correctness), distributional (output diversity collapses to stereotypical modes), and metacognitive (the model fails to assess its own competence). |
Tiancheng Hu; Caiqi Zhang; Dirk Hovy; Nigel Collier; |
| 359 | The Dead Salmons of AI Interpretability: The Need for A Statistical Inference Perspective Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Astonishingly, standard analyses of the time reported brain regions \textit{predictive} of social emotions. |
Maxime Méloux; Giada Dirupo; François Portet; Maxime Peyrard; |
| 360 | Validating Causal Abstraction Metrics on Simulated Complex Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce a benchmark of ten complex systems spanning discrete and continuous state spaces and static and dynamical regimes, each equipped with consensual ground-truth causal explanations and invalid contrastive conditions. |
Maxime Méloux; Tiago Pimentel; François Portet; Maxime Peyrard; |
| 361 | MedKIT: Evaluating Knowledge Integration and Generalization in Large Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our benchmark __MedKIT__ (Medical Knowledge Integration and Transfer) provides a granular evaluation of how models integrate and apply knowledge under realistic sequences of clinical updates. |
Lukas Thede; Yash Kumar; David Chen; Danielle Bitterman; Matthias Bethge; Tom Hartvigsen; Zeynep Akata; |
| 362 | InfoSFT: Learn More and Forget Less with Information-Aware Token Weighting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose InfoSFT, a principled weighting scheme for the SFT objective that concentrates learning signals on maximally informative, medium-confidence tokens—those neither overly familiar to the base model nor too unlikely to cause instability. |
Mahdi Sabbaghi; George J. Pappas; Adel Javanmard; Hamed Hassani; |
| 363 | Robust Policy Optimization to Prevent Catastrophic Forgetting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Fine-tuning Robust Policy Optimization (FRPO), a robust RLHF framework that optimizes reward not only at the current policy, but across a KL-bounded neighborhood of policies reachable by downstream adaptation. |
Mahdi Sabbaghi; George J. Pappas; Adel Javanmard; Hamed Hassani; |
| 364 | ExploitGym: Can AI Agents Turn Security Vulnerabilities Into Real Attacks? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Despite its importance and diagnostic value, exploitation remains under-evaluated. To address this gap, we introduce ExploitGym, a large-scale, diverse, realistic benchmark on the exploitation capabilities of AI agents. |
Zhun Wang; Nico Schiller; Hongwei Li; Srijiith S Narayana; Milad Nasr; Nicholas Carlini; Xiangyu Qi; Eric Wallace; Elie Bursztein; Luca Invernizzi; Kurt Thomas; Yan Shoshitaishvili; Wenbo Guo; Jingxuan He; Thorsten Holz; Dawn Song; |
| 365 | EverAnimate: Minute-Scale Human Animation Via Latent Flow Restoration Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose **EverAnimate**, an efficient post-training method for long-horizon animated video generation that preserves visual quality and character identity. |
WUYANG LI; Yang Gao; Mariam Hassan; Lan Feng; Wentao Pan; Po-Chien Luan; Alexandre Alahi; |
| 366 | S-EDL: Eliciting Self-Evidence from Sequence Likelihoods for Semantic Calibration of LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Prior calibration techniques target either fixed-set classification confidence or the binary correctness of individual strings, making them inherently ill-suited for the dynamic, multi-string nature of open-ended semantic outcomes. To bridge this gap, we introduce S-EDL, a novel evidential method that natively optimizes LLMs for semantic calibration. |
Yawei Li; Jiazheng Li; David Rügamer; Bernd Bischl; Yulan He; Mina Rezaei; |
| 367 | Learning to Persuade Exposes How Easily LLMs Abandon Correct Beliefs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Yet we show that this requirement is far from met: a single targeted persuasive argument is enough to collapse model accuracy to near zero, even when the argument is factually false. We formalize this threat as adversarial persuasion and introduce an adversarial reinforcement learning framework that trains persuader agents to change a target model’s answer in a single interaction. |
Nimet Beyza Bozdag; Emre Can Acikgoz; Gokhan Tur; Dilek Hakkani-Tur; |
| 368 | RadOmni: Advancing Foundation Model for Non-contrast CT with Omni Radiology Knowledge Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, existing pretraining paradigms address this challenge only partially, each introducing complementary but still limited constraints on NCCT representations. To address this limitation, we propose RadOmni, a unified omni-pretraining framework that injects multimodal radiology knowledge into representation learning for non-contrast CT. |
Fenghe Tang; Weiwei Cao; Wenxin (Wendy) Ma; Kai Cao; Huanhuan Liu; Ling Zhang; S. Kevin Zhou; Jianpeng Zhang; |
| 369 | MultiTalk: Scaling Full-Duplex Speech Models to Long, Multi-Party, Bilingual Conversation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we extend the Moshi paradigm along long-horizon and multi-party axes simultaneously, in both English and Chinese, with three contributions.First, we release an open data engine and a 57.6 k-hour corpus for long, multi-party, bilingual full-duplex dialogue. |
Ke Wang; Houxing Ren; Zimu Lu; Yunqiao Yang; ZHUOFAN ZONG; Mingjie Zhan; Hongsheng Li; |
| 370 | Trimming The Long-Tail of Visual World Modeling Evaluation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This raises a central question: Do world models internalize and generalize the physical principles? In this work, we introduce Tailor-Bench, a benchmark to systemically evaluate this question through the lens of visual world modeling on tool-using interactions. |
Bingxuan Li; Yining Hong; Cheng Qian; Hyeonjeong Ha; Jiateng Liu; Zhenhailong Wang; Yue Guo; Yunzhu Li; Heng Ji; |
| 371 | Correct But Unselectable: The Hidden Interface Tax in Multi-Candidate Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Compatibility Aware Candidate Construction (CACC), an interface layer between the candidate source and the fixed selector that repairs fragments, aligns answer modes, and removes selector-confusing noise without retraining either component. |
Qianli Ma; Zhiqing Tang; Fanshuai Meng; Zhi Yao; Hanshuai Cui; Weijia Jia; |
| 372 | Representation Forcing for Bottleneck-Free Unified Multimodal Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Naively removing it introduces a quality gap, as the model must learn both high-level structure and low-level details from raw pixels. In this paper, we propose Representation Forcing (RF), a technique that closes this gap by making representation prediction a native capability of the decoder. |
Yuqing Wang; Zhijie Lin; Ceyuan Yang; Yang Zhao; Fei Xiao; Hao He; Qi Zhao; Zihan Ding; Fu-Yun Wang; Shuai Wang; Youliang Zhang; Haoqi Fan; Xihui Liu; |
| 373 | From Infrastructure to Interface, The AI Value Chain Drives LLM Homogenization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, most approaches address outcome homogenization exclusively through a technical lens. In this position paper, we argue that outcome homogenization is best understood as a value-chain phenomenon shaped by stakeholders, governance structures, resource allocation, model development, and organizational practices. |
Khaoula Chehbouni; Cléa Chataigner; Prakhar Ganesh; Pablo Piantanida; Jackie CK Cheung; Golnoosh Farnadi; |
| 374 | StreamPI: Streaming Multimodal Temporal Modeling for Vision-Language-Action Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose StreamPI, a streaming multimodal temporal modeling framework that equips single-frame VLA with temporal reasoning capability without introducing any additional parameters. |
Zhe Liu; Jinghua Hou; Yuxiang Lu; Zhenya YANG; Xianzhe Fan; Junwei Luo; Junyi Li; Ruihua Han; Zhi Hou; Hengshuang Zhao; |
| 375 | MDMR-Bench: A Multi-Dimensional Benchmark for Multi-Reference Image Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, existing benchmarks for multi-reference image generation rely on limited evaluation dimensions and fail to capture the diverse challenges of the task. To address this limitation, we propose **MDMR-Bench**, a **M**ulti-**D**imensional benchmark for **M**ulti-**R**eference image generation. |
Bowen Zheng; Jintao Lin; Chang-Heng Yi; HAORAN YANG; Binxiao Huang; Chenyang Lei; Rui Liu; Xihui Liu; |
| 376 | From Squeezing to Grounding: Visual Guided DPO for Multimodal Hallucination Mitigation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Visual Guided Direct Preference Optimization (VGDPO), which relaxes negative updates on rejected responses according to multimodal squeezing risk. |
Wenqi Liu; Hongxin He; Yunxiao Wang; Xuemeng Song; Yupeng Hu; Yinwei Wei; |
| 377 | EO-WM: A Physically Informed World Model for Probabilistic Earth Observation Forecasting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we view this task as a partially observed, weather-driven world modeling problem, in which weather acts as a conditioning signal, while forecasting remains uncertain due to sparse observations and unobserved land-surface states.To evaluate weather-response behavior beyond standard metrics, we introduce two diagnostic benchmarks: an Extreme Summer Benchmark for severity-aware prediction of vegetation degradation under extreme weather, and a Seasonal Matched-Pair Benchmark for testing response fidelity under changed weather forcing. |
Junwei Luo; Shuai Yuan; Zhenya YANG; Yansheng Li; Zhe Liu; Hengshuang Zhao; |
| 378 | Harder The Task, Sparser The Representation: Sparsity As A Learning Signature of Capability in LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we investigate how the internal representations of Large Language Models (LLMs) change with inputs of increasing difficulty. |
Mingyu Jin; Yutong Yin; Jingcheng Niu; Qingcheng Zeng; Wujiang Xu; Xinyuan Song; Wei Cheng; Mengnan Du; Zhaoran Wang; Tianlong Chen; Dimitris Metaxas; |
| 379 | FlowR2A: Learning Reward-to-Action Distribution for Multimodal Driving Planning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Multimodal driving planning faces a long-standing tension between two paradigms: scoring-based methods benefit from dense reward supervision but are confined to a fixed action vocabulary, while anchor-based methods generate proposals dynamically yet suffer from sparse supervision constrained to a single ground-truth trajectory. In this work, we propose FlowR2A, which resolves this tension by reframing simulation-based rewards from discriminative targets into generative conditions. |
Xirui Li; Zhe Liu; Xiaoqing Ye; Wenhua Han; Yifeng Pan; Junyu han; Hengshuang Zhao; |
| 380 | Do Semantic Distance Tests Actually Predict Creativity in Large Language Models? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We score each test on two criteria (*validity* and *specificity*), and derive a theoretical limit for the maximum attainable specificity and validity a test can achieve. |
Samuel Schapiro; Alexi Gladstone; Jonah Black; Heng Ji; |
| 381 | DDx-TRACE: A Benchmark for Medical Diagnostic Trajectories in VLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce DDx-TRACE, a physician-adjudicated benchmark for multimodal neuroradiology that evaluates diagnostic trajectories under hidden evidence over 211 challenging cases. |
Jiazhen Pan; Weixiang Shen; Jun Li; Julian Canisius; Felix Bitzer; Paula Roßmüller; Jiancheng Yang; Virginie Kreutzinger; Daniel Rueckert; Benedikt Wiestler; |
| 382 | Unlocking LLM Creativity in Science Through Analogical Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We evaluate LLMs on the task of open-ended solution generation and quantify their tendency to mode collapse into low-diversity generations. To mitigate this mode collapse, we introduce analogical reasoning (AR) as a new approach to solution generation. |
Andrew Shen; Shaul Druckmann; James Zou; |
| 383 | CineOrchestra: Unified Entity-Centric Conditioning for Cinematic Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present $\textit{CineOrchestra}$, a unified video diffusion model that controls subjects, events, cameras, and shot transitions simultaneously. |
Sharath Girish; Tsai-Shien Chen; Zhikang Dong; Mukesh Singhal; Hao Chen; Sergey Tulyakov; Aliaksandr Siarohin; |
| 384 | FLARE: Diffusion for Hybrid Language Model Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we present FLARE, a systematic recipe for converting hybrid AR LLMs into capable, real-time-fast dLLMs under a practical training budget. |
Yuchen Zhu; Jing Shi; Chongjian GE; Hao Tan; Yiran Xu; Wanrong Zhu; Jason Kuen; Ryan Rossi; Koustava Goswami; Rajiv Jain; Yongxin Chen; Molei Tao; Jiuxiang Gu; |
| 385 | LinuxArena: A Control Setting for AI Agents in Live Production Software Environments Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: The field of AI control develops control protocols that prevent such harm without restricting the agent’s ability to do useful work. We introduce \textbf{LinuxArena}, a control setting for testing such protocols, where agents operate directly on live, multi-service production environments. |
Tyler Tracy; Ram Potham; Nikolas Kuhn; Myles Heller; Anshul Khandelwal; Cody Rushing; Henri Lemoine; Miguel Brandão; Tomáš Turlík; Adam Hanson; Josh Hills; Amy D Ngo; Ram Rachum; Nik Mitchell; Falko Galperin; Oscar Sykes; Pip Arnott; Samuel P Lima; Carlos R Giudice; Matt Goldwater; Daniel J Popp; Drew de Wet; Ruben Castaing; Qi Guo; Douw Marx; Benjamin Shaffrey; Justin Shenk; Martin Milbradt; Hannah Meagher; Shaheen Ahmed-Chowdhury; Daniel O’Connell; Christopher W Canal; Buck Shlegeris; Aryan Bhatt; |
| 386 | Proteus: Incremental Memory Activation for Long-Context Sequence Modeling Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study a new paradigm of incremental memory activation, where the effective capacity of memory is controlled and progressively expanded as the context grows. |
Reza Bayat; Ali Behrouz; Vahab Mirrokni; Aaron Courville; |
| 387 | Convergence of Near-Linear Width ReLU Networks with Unbalanced Initialization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Linear-width convergence for two-layer ReLU networks is known in random-feature or conjugate kernel (CK) regimes whose analyses exploit nearly fixed hidden features. |
Zhenghao Xu; Rachel Ward; Molei Tao; Tuo Zhao; |
| 388 | BRAVO: Bridge Matching for Autoregressive Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To this end, we introduce BRAVO, a Brownian bridge matching framework for autoregressive video generation. |
Shihua Zhang; Zhenxiong Tan; Xinchao Wang; |
| 389 | Breaking Feature Collapse in Self-supervised Time Series Encoder Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce \texttt{OTISv2}, which suppresses the collapse by complementing the masked-reconstruction objective with two representation-space terms — a self-distillation loss against an exponential-moving-average teacher and a KoLeo regulariser — and replaces \texttt{OTIS}’s per-domain variate embeddings with register tokens to remove the label dependency. |
Özgün Turgut; Daniel Rueckert; |
| 390 | OTIS: Learning High-Quality Time Series Features With Tiny Encoders Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce OTIS, an open time series encoder that yields high-quality time series features for downstream deployment on any system, including resource-constrained wearables and industrial sensors. |
Özgün Turgut; Philip Müller; Martin Menten; Daniel Rueckert; |
| 391 | Hyperagents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce \textbf{hyperagents}, self-referential agents that integrate a task agent (which solves the target task) and a meta agent (which modifies itself and the task agent) into a single editable program. |
Jenny Zhang; Bingchen Zhao; Wannan Yang; Jakob Foerster; Jeff Clune; Minqi Jiang; Sam Devlin; Tatiana Shavrina; |
| 392 | Memory Grafting: Scaling Language Model Pre-training Via Offline Conditional Memory Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To this end, we propose Memory Grafting, a representation-level transfer paradigm that extracts structured memories from a large frozen donor model and injects them into a compact student throughout training. |
Runxi Cheng; Yuchen Guan; Yongxian Wei; Qianpu Sun; Qixiu Li; Sinan Du; Feng Xiong; Chun Yuan; Yan Lu; Yeyun Gong; |
| 393 | Seeking The Unfamiliar But Memorable: Conceptual Creativity As Meta-Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose an operational definition of creativity as the generation of concepts that are initially unfamiliar yet rapidly learnable to an adaptive observer, and formalize it as a bilevel optimization between a Creator that generates and an Appraiser that adapts: the Appraiser’s improvement under a brief inner-loop adaptation provides the reward signal that the Creator maximizes. |
Mengye Ren; |
| 394 | Bernini: Latent Semantic Planning for Video Diffusion Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We argue that these two families can be unified through a simple division of labor: MLLMs perform semantic planning, while diffusion models render pixels from high-level semantic guidance and low-level visual features. Building on this idea, we propose \textbf{Bernini}, a unified framework for video generation and editing. |
Chenchen Liu; Junyi Chen; Lei Li; Lu Chi; Mingzhen Sun; Zhuoying Li; Yi Fu; Ruoyu Guo; Yiheng Wu; Ge Bai; Zehuan Yuan; |
| 395 | EvalAwareBench: A Benchmark for Measuring Evaluation Awareness in Frontier Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce EvalAwareBench, an open pipeline and benchmark for measuring evaluation awareness that works with any Inspect-compatible evaluation, allowing practitioners to test against current and future benchmarks. |
Xinning Li; Kemunto Ochwang’i; Ram Bharadwaj Aryasomayajula; Alexandra Souly; Robert Kirk; |
| 396 | Large Language Models Suffer from A Curse of Ambiguity Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we identify a curse of ambiguity: in large language models, and more broadly in all neural networks that produce discrete probability distributions, the more ambiguous a next-token distribution is, the harder it is to learn accurately. |
Nicolas Zucchet; Hyun Dong Lee; Scott Linderman; |
| 397 | Are Text-to-Image Models Inductivist Turkeys? A Counterfactual Benchmark for Causal Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We evaluate both open-source and closed-source T2I models using a Vision Language Model(VLM)-based evaluator (CF-Eval), and introduce two metrics: Prior Resistance Rate (PRR), which measures a model’s ability to overcome entrenched real-world priors, and Reasoning Retention Rate (RRR), which assesses whether models can maintain reasoning-dependent counterfactual generation without explicit visual cues. |
jiayi lei; Yuandong Pu; Xingyu Han; Rongpeng Zhu; Jing Xu; Jinyao Wang; Zijian Zhou; Bin Fu; Yuewen Cao; Yihao Liu; Hongsheng Li; |
| 398 | Designing Cell-Type-Specific Regulatory DNA with Guided Discrete Diffusion Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Here, we introduce DNA-CRAFT, a method for designing regulatory DNA with high predicted cell-type-specific activity while preserving naturalness. |
Animesh Awasthi; Martin Stoll; Raphael Bednarsky; Moritz Schaefer; Christoph Bock; |
| 399 | WorldPrism: 3D Consistency for Video World Models Via Bidirectional Cross-Space Verification Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose WorldPrism, a post-training framework that improves 3D consistency via reinforcement learning without modifying the base model architecture. |
Hengyu Liu; Jiahao Lu; Ruijie Zhu; Sixiao Zheng; Yuan Liu; Wenbo HU; Ying Shan; |
| 400 | OpenCoF: Learning to Reason Through Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, existing video generators are primarily trained on general video corpora, still lacking diverse supervision and dedicated designs for CoF reasoning. To address this gap, we introduce OpenCoF, a framework comprising the OpenCoF-17K Dataset, a reasoning video dataset spanning 11 task families, and Wan-CoF, a fine-tuned video model for studying whether diverse temporal supervision improves CoF behavior. |
Xinyan Chen; Renrui Zhang; Ziyu Guo; Dongzhi JIANG; Hongsheng Li; |
| 401 | A Minimal Interpretable Architecture for Zero-Shot Reconstruction of Dynamical Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Such an understanding could help to strip down overladen FM architectures to their bare essence and expose the minimal requirements for in-context learning in the DS domain. Toward this goal, here we iteratively reduce a recent powerful SOTA model for DS reconstruction, DynaMix (Hemmer & Durstewitz, 2025), to a minimal interpretable two-parameter form, which we call *DynaBase*. |
Christoph Jürgen Hemmer; Florian Plaswig; Daniel Durstewitz; |
| 402 | Effective Multi-sensor Conditioning for Street-view Novel-view Synthesis Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce StreetNVS, a video diffusion framework that jointly conditions on all three signals through a Reference-Enhanced Camera Attention module based on relative ray-level positional encoding, trained with a two-stage curriculum that gradually exposes the model to increasingly sparse LiDAR. |
Zhengfei Kuang; Adam Sun; Liyuan Zhu; Tong Wu; Shengqu Cai; Jonathan Tremblay; Iro Armeni; Ehsan Adeli; Lior Yariv; Gordon Wetzstein; |
| 403 | Working with AI: Measuring The Applicability of Generative AI to Occupations Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We use an LLM-based pipeline to classify the O*NET work activities assisted or performed by AI in each conversation. |
Kiran Tomlinson; Sonia Jaffe; Will Wang; Scott Counts; Siddharth Suri; |
| 404 | Screening Lipid Nanoparticles Through Structure-Ratio Alignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose STRATA, a framework that models molecule interaction between LNP components, which is known to contribute to LNP transfection efficiency. |
Yoonho Lee; Yunhak Oh; Hoyoung Choi; Chanyoung Park; |
| 405 | An Embarrassingly Simple Graph Heuristic Reveals Shortcut-Solvable Benchmarks for Sequential Recommendation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Rather than treating datasets as interchangeable leaderboards, we argue for more careful dataset selection and dataset-level diagnostic analysis when using benchmarks to support claims about the benefits of new recommendation models. |
Haoyu Han; Li Ma; Hanbing Wang; Bingheng Li; Daochen Zha; Chun How Tan; Huiji Gao; Xin Liu; Stephanie Moyerman; Sanjeev Katariya; Hui Liu; Jiliang Tang; |
| 406 | Embodied-R1.5: Evolving Physical Intelligence Via Embodied Foundation Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce Embodied-R1.5, a unified Embodied Foundation Model (EFM) that integrates comprehensive embodied reasoning capabilities, spanning embodied cognition, task planning, correction, and pointing, within a single architecture toward general physical intelligence. |
Yifu Yuan; Yaoting Huang; Xianze Yao; Shuoheng Zhang; Linqi Han; Yutong Li; Pengyi Li; Jiangeng Sun; Wenting Jia; Yucheng Hu; YuHao Liu; Ruihao Liao; Qiyu Wu; Yuxiao Li; zhao zhang; Zibin Dong; Fei Ni; YAN ZHENG; Shuyang Gu; Yi Ma; Hongyao Tang; Han Hu; Jianye Hao; |
| 407 | TMPO: Trajectory Matching Policy Optimization for Diverse and Efficient Diffusion Alignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In contrast, we propose Trajectory Matching Policy Optimization (TMPO), which replaces scalar reward maximization with trajectory-level reward distribution matching. |
Jiaming Li; Chenyu Zhu; Zhiyuan Ma; Nanxi Yi; Youjun Bao; Li Sun; Quanying Lv; Xiang Fang; Daizong Liu; Jianjun Li; Kun He; Bowen Zhou; |
| 408 | Generalizable Physics Simulation Through Compositional Energy Minimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce Compositional Potential Minimization (CPM), a framework that casts simulation as an energy minimization procedure over a composition of learned force potentials. |
Alexandru Oarga; Yilun Du; |
| 409 | Permutation Sensitivity in T-SVD-based Multi-view Clustering Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Extensive experiments demonstrate the significant performance variability of existing t-SVD-based methods under sample permutations, whereas our approaches substantially improve their permutation robustness while maintaining competitive clustering performance. |
Jintian Ji; Yanjun Zhang; He Zhang; Leo Yu Zhang; Cong Wang; Shirui Pan; |
| 410 | PRECISE: SDE-Consistent Stochastic Sampling for RL Post-Training of Flow-Matching Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Turning to the discretization challenge, we use a toy example to show that existing samplers can deviate from the flow-matching process, either by introducing excessive discretization noise or by relying on heuristic rules that do not guarantee convergence to the data distribution. To address these issues, we propose \textsc{Precise}, a new stochastic sampler that balances effective exploration with stability. |
Bo Peng; Tao Huang; Weijie Kong; Junzhe Li; Yue Wu; Qi Tian; Jiangfeng Xiong; Jian-Wei Zhang; Liefeng Bo; Zhao Zhong; |
| 411 | Statistical Estimation of Adversarial Risk in Large Language Models Under Best-of-N Sampling Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a scaling-aware Best-of-N estimation of risk, SABER, for modeling jailbreak vulnerability under Best-of-N sampling. |
Mingqian Feng; Xiaodong Liu; Weiwei Yang; Chenliang Xu; Christopher White; Jianfeng Gao; |
| 412 | Unified Panoramic Geometry Estimation Via Multi-View Foundation Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce PaGeR (Panoramic Geometry Reconstruction), a framework to lift powerful 3D foundation models, designed for perspective imagery, to the panorama domain. |
Vukasin Bozic; Isidora Slavkovic; Dominik Narnhofer; Nando Metzger; Denis Rozumny; Konrad Schindler; Nikolai Kalischek; |
| 413 | Beyond Semantic Similarity: Rethinking Retrieval for Agentic Search Via Direct Corpus Interaction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While efficient, this design becomes a bottleneck for agentic search: exact lexical matching, conjunctions of weak clues, local context verification, and multi-step hypothesis refinement are difficult to realize with a conventional off-the-shelf retriever, and evidence filtered out early cannot be recovered downstream. This problem becomes particularly acute when agents must iteratively discover intermediate entities, combine partial evidence, and revise their search strategy. |
Zhuofeng Li; Haoxiang Zhang; Cong Wei; Pan Lu; Ping Nie; Yuyang Bai; Shangbin Feng; Hangxiao Zhu; Ming Zhong; Yuyu Zhang; Jianwen Xie; Yejin Choi; James Zou; Jiawei Han; Wenhu Chen; Jimmy Lin; Dongfu Jiang; Yu Zhang; |
| 414 | RxnOptBench: Benchmarking LLMs for Reaction-Condition Optimization in Organic Methodology Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce RxnOptBench, a benchmark whose every option and precedent is a real wet-lab entry mined from the optimization tables of organic-methodology papers published in 2025, graded by a continuous relative score that fuses yield with stereoselectivity (ee, dr, rr) so near-correct answers are not collapsed to zero, and equipped with a paired precedents-vs-no-precedents design that isolates in-context use of literature evidence from parametric memorization.We release the benchmark, the human-verified extraction pipeline, and the inference and evaluation code. |
Lingli Ge; Junyuan Gao; Jiahe Song; Jiaxing Sun; Yubin Wang; Boyu Zhu; Haote Yang; Jingchao Wang; Lixin Ma; Jiang Wu; Yuqiang Li; Conghui He; |
| 415 | Rebellious Student: Reversing Teacher Signals for Reasoning Exploration with Self-Distilled RLVR Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Therefore, we propose reading the original self-distillation signal in reverse: when the student succeeds along a path the teacher would not have predicted, these tokens reflect its self-driven reasoning. Building on this, we propose RLRT (RLVR with Reversed Teacher), which augments GRPO by reinforcing these tokens on correct rollouts. |
Jeonghye Kim; Jiwon Jeon; Dongsheng Li; Yuqing Yang; |
| 416 | Does Seeing More Mean Knowing More? Mono-Anchored Advantage Normalization for Multi-Source Visual Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Therefore, they struggle to effectively model dynamic interaction when integrating multiple sources, particularly when they differ significantly in physical properties and semantics, \eg, infrared and depth, leading to inferior performance to mono-source reasoning when a certain source holds the dominant signal. To address this issue, we propose MARS, a novel mono-anchored multi-source reasoning framework that models each visual modality as an independent information source. |
Fanhu Zeng; Zhicong Luo; Zefan Wang; Li You; Chi Chen; Maosong Sun; |
| 417 | Spectral Progressive Diffusion for Efficient Image and Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Spectral Progressive Diffusion, a general framework that progressively grows resolution along the denoising trajectory of pretrained diffusion models. |
Howard Xiao; Brian Chao; Lior Yariv; Gordon Wetzstein; |
| 418 | Social Interaction Breaks Replicate Independence in Controlled-Clone LLM Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Multi-agent LLM evaluations and large-scale social simulations often treat multiple copies of the same model as comparable replicates after they interact. We test this assumption with SocCogEval, a controlled-clone assay for Social Cognitive Transfer (SCT), and find that it fails. |
Yandan Zheng; Haoran Luo; Anh Tuan Luu; |
| 419 | Forecasting Downstream Performance of LLMs With Proxy Metrics Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Instead, we propose to construct proxy metrics by aggregating token-level statistics, such as entropy, top-k accuracy, and expert token rank, from a candidate model’s next token distribution over expert-written solutions. |
Arkil Patel; Siva Reddy; Marius Mosbach; Dzmitry Bahdanau; |
| 420 | Mixture of Layers: Dynamic Layer Routing for Visual Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose **Mixture of Layers (MoL)**, an instruction-conditioned layer routing approach that dynamically aggregates query-relevant latent representations from intermediate vision encoder layers. |
Jeonghwan Kim; Sofia Stoica; Jiwan Chung; Ansel Blume; Hyeonjeong Ha; Zhenhailong Wang; Xin Dong; Heng Ji; |
| 421 | OmniMemBench: Towards Scalable Evaluation of Long-Term Omni-Modal Agent Memory Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce OmniMemBench, to our knowledge the first long-term memory benchmark that spans image, audio, and video in multi-session dialogue. |
Junhan Shi; Qiyi Wang; Xiongwei Wu; Ming Ma; Baoqi Pei; Zhenning Shi; Yong Jiang; Qing Li; Weigao Sun; Yiran Zhong; Steven Hoi; |
| 422 | Reasoning Gestures: LLM-Inferred Communicative Functions for Co-Speech Gesture Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Most co-speech gesture generation methods, however, condition primarily on low-level cues (speech audio or transcripts) and do not explicitly represent these higher-level functions, leading to motions that synchronize with speech but remain under-specified in communicative meaning. To address this gap, we introduce \textbf{Reasoning Gestures}, a reasoning-grounded framework that conditions generation on interpretable communicative-function descriptions. |
Pinxin Liu; Haiyang Liu; Yunlong Tang; Luchuan Song; |
| 423 | Entangled Schrödinger Bridge Matching Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, many systems, including biomolecular systems and heterogeneous cell populations, undergo *dynamic* interactions that evolve over their trajectory and cannot be captured through static snapshots. To close this gap, we introduce **Entangled Schrödinger Bridge Matching (EntangledSBM)**, a framework that learns the first- and second-order stochastic dynamics of interacting, multi-particle systems where the direction and magnitude of each particle’s path depend dynamically on the paths of the other particles. |
Sophia Tang; Yinuo Zhang; Pranam Chatterjee; |
| 424 | Expanding Flow Maps Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Here, we introduce **Expanding Generative Flows** (EFlows), which define flows between distributions of *increasing dimensionality* through an expanding interpolant built from *augmented distributions of conditional noise*. |
Sophia Tang; Pranam Chatterjee; |
| 425 | Beyond Task Success: Probing Cognitive Primitives in Web Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Because such tasks are not designed to require specific primitives, observed behaviors do not reliably reflect true capability demands. This raises a key question: Can we design realistic tasks where success provably depends on a chosen cognitive primitive? |
Xunjian Yin; Tianchen Guan; Jinao Wang; Weili Cao; Royce Cheng-Yue; Daisy Lin; Keagan Long; Kyle Wong; Bhuwan Dhingra; XJ Wang; Shuyan Zhou; |
| 426 | SRL-MPC: Shape-Aware Reinforcement Learned Model Predictive Control Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Traditional approaches often assume homogeneous robots, sparse workspaces, simplified geometry, offline computation, or handcrafted parameters to make the problem tractable, which limits their deployment in dense crowd scenarios. Toward this end, we propose Shape-Aware Reinforcement Learned Model Predictive Control (SRL-MPC), a method for safe, efficient, and adaptive navigation in crowds with heterogeneous shapes without geometry simplification. |
Ruihua Han; Rui Gao; Zhe Liu; Xinyi WANG; Chang Chen; Shuai Wang; Qi Hao; Jia Pan; Hengshuang Zhao; |
| 427 | ViTeX-Bench: Benchmarking High-Fidelity Video Scene Text Editing Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing resources do not provide paired edits over real-world videos, current protocols rarely measure whether the edited region reads as the target string over time, and there is no large-scale open-source benchmark for systematic comparison. To close these gaps, we introduce ViTeX-Bench, a comprehensive benchmark suite for high-fidelity and temporally consistent video scene text editing. |
Xinghao Chen; Xiangbo Gao; Jiongze Yu; Yuheng Wu; Zhengzhong Tu; |
| 428 | Breaking The Quality–Privacy Tradeoff in Tabular Data Generation Via In-Context Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This tradeoff arises because small training sets make it difficult for dataset-specific generative models to distinguish generalizable structure from sample-specific patterns. To address this, we propose DiffICL, which formulates tabular data generation as an in-context learning problem. |
Xinyan Han; yan Lu; Xiaoyu Lin; Yuanyuan Jiang; Yuanrui Wang; Xuanyue Li; Wenchao Zou; Xingxuan Zhang; |
| 429 | EvoMM: Reinforced Self-Evolving Multimodal Agentic Memory Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Closing this loop with reinforcement learning additionally suffers from credit assignment difficulties over long, multimodal action trajectories. To address these challenges, we propose EvoMM, a reinforced self-evolving multimodal agentic memory framework that unifies memory construction, retrieval, and evolution within a single agentic loop. |
Tong Zhao; Chenghao Zhang; Yucheng Tian; Yuyang Hu; Yutao Zhu; Zhicheng Dou; |
| 430 | Articraft: An Agentic System for Scalable Articulated 3D Asset Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: A bottleneck in learning to understand articulated 3D objects is the lack of large and diverse datasets. In this paper, we propose to leverage large language models (LLMs) to close this gap and generate articulated assets at scale. |
Matt Zhou; Ruining Li; Xiaoyang Lyu; Zhaomou Song; Zhening Huang; Chuanxia Zheng; Christian Rupprecht; Andrea Vedaldi; Shangzhe Wu; |
| 431 | AVO: Agentic Variation Operators for Autonomous Evolutionary Search Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We evaluate AVO on attention, among the most aggressively optimized kernel targets in AI, on NVIDIA Blackwell (B200) GPUs. |
Terry Chen; Zhifan Ye; Bing Xu; Zihao Ye; Timmy Liu; Ali Hassani; Tianqi Chen; Andrew Kerr; Haicheng Wu; Vedaanta Agarwalla; Fengzhe Zhou; Shangda Li; Yang Xu; Yu-Jung Chen; Hanfeng Chen; Aditya Kane; Ronny Krashinsky; Ming-Yu Liu; Vinod Grover; Luis Ceze; Roger Bringmann; John Tran; Wei Liu; Feng Xie; Michael Lightstone; Humphrey Shi; |
| 432 | Political Neutrality As Balanced Approval: A Large-Scale Human Evaluation of AI Responses Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Here, we propose a new definition of AI political neutrality and design a large-scale user study to test it, releasing a new dataset PARETO with 6,696 participants and 187,488 evaluations of AI responses.We construct a benchmark of controversial US issues, with prompts sourced from politically charged questions on Reddit and responses from frontier AI models, and recruit human participants to rate AI responses.This work introduces a rigorous definition and benchmark of AI political neutrality, and a dataset to measure progress toward it. |
Jonathan Stray; David Yang; Steven Luo; Miu N Takagi; Serina Chang; |
| 433 | Sparse Koopman Autoencoders Identify Local Dynamical Regimes in Multibasin Systems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Across a range of procedurally generated multibasin systems and chaotic flows, we show that SKAEs have superior forecasting performance compared to dense-latent KAEs. |
Aidan Li; Uday Kiran Reddy Tadipatri; Mahan Fathi; Sarath Chandar; Ross Goroshin; |
| 434 | TALES: Text-Adventure Learning Environment Suite Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, current implementations in popular benchmarks obscure performance through the leakage of privileged information or through scaffolding modifications to the tasks themselves. To solve this, we introduce \ours, a diverse collection of synthetic and human-written text-adventure games in their canonical forms, designed to evaluate baseline agentic competencies in progressively more challenging environments. |
Christopher Z. Cui; Xingdi Yuan; Ziang Xiao; Prithviraj Ammanabrolu; Marc-Alexandre Côté; |
| 435 | Continuous Latent Diffusion Language Model Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose Cola DLM, a hierarchical continuous latent diffusion language model that learns a stable Text VAE, models a global semantic prior in latent space with a block-causal DiT, and decodes text conditionally. |
Hongcan Guo; Qinyu Zhao; Yian Zhao; Shen Nie; Rui Zhu; Qiushan Guo; Feng Wang; Tao Yang; Hengshuang Zhao; Guoqiang Wei; Yan Zeng; |
| 436 | TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce TerminalWorld, a scalable data engine that automatically reverse-engineers high-fidelity evaluation tasks from in-the-wild terminal recordings. |
Zhaoyang Chu; Jiarui Hu; Xingyu Jiang; Pengyu Zou; Han Li; Chao Peng; Peter O’Hearn; Earl Barr; Mark Harman; Federica Sarro; He Ye; |
| 437 | Mitigating Overthinking in Large Reasoning Language Models Via Reasoning Path Deviation Monitoring Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We observe that when an LRLM’s reasoning path deviates into meaningless or even harmful wandering, this deviation is accompanied by an anomalous surge of high-entropy transition tokens. Building on this insight, we propose RPDI-EE, a training-free early-exit method that monitors the Reasoning Path Deviation Index (RPDI), defined as the ratio of local to global average entropy, which serves as a proxy for reasoning path deviation. |
Weixin Guan; Liang Li; Jiapeng Liu; Bing Li; Peng Fu; Chengyang Fang; Xiaoshuai Hao; Can Ma; Weiping Wang; |
| 438 | PREPING: Building Agent Memory Without Tasks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we study pre-task memory construction: whether an agent can build procedural memory before observing any target-environment tasks, using only self-generated synthetic practice. |
Yumin Choi; Sangwoo Park; Minki Kang; Jinheon Baek; Sung Ju Hwang; |
| 439 | Persistent Planar Memory for Video World Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Plane primitives offer a natural candidate: a compact set of oriented planes captures the dominant structural elements (*e.g.,* walls, floors, facades) of indoor and outdoor scenes, while requiring only 7 parameters per primitive. We present **PPMem**, a video world model that adopts plane primitives as its persistent 3D memory. |
Yuze He; Bin Tan; Zelin Gao; Yujun Shen; Yong-jin Liu; Nan Xue; Yinghao Xu; |
| 440 | Useful Memories Become Faulty When Continuously Updated By LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Recent work on agentic-memory systems has explored a practical route to continuously improving LLM agents after deployment without parameter updates: a textual abstraction is derived from each history and stored in memory, then continuously updated with more interactions. Yet we show that such textual abstractions produced by today’s LLMs are often faulty, even when derived from useful experiences. |
Dylan Zhang; Yanshan Lin; Zhengkun Wu; Yihang Sun; Bingxuan Li; Dianqi Li; Hao Peng; |
| 441 | OfficeInstruct: A Dataset for Tool-Augmented Office Agents on Long-Trajectory Generative Tasks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To ensure data quality, we propose a graph-based evidence evaluation method that models interaction histories as directed graphs, filtering out redundant tool cycles, execution errors, and hallucinations. |
Weikai Xu; Zhizheng Jiang; Lushuo Jiang; Kun Huang; Yuxuan Liu; Pengzhi Gao; Wei Liu; Jian Luan; Xiaolin Hu; Bo An; |
| 442 | Offloading Score: Measuring AI Reliance Through Counterfactual Workflows Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Here, we introduce offloading score, a measure of reliance that quantifies the fraction of cognitive effort offloaded to an AI tool. |
Vishakh Padmakumar; Lujain Ibrahim; Zora Wang; Jennifer Wang; Q.Vera Liao; Diyi Yang; |
| 443 | ToS: Tree-of-Skill Reinforcement Learning Driven By Behavior Tree for Long-Horizon Manipulation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This paper proposes Tree-of-Skill (ToS), the first neural-symbolic framework for dynamic skill chaining in BT-driven Reinforcement Learning (RL). |
Xinglin Chen; Yishuai Cai; Yunxin Mao; Minglong Li; Kefei Zhu; Wenjing Yang; Yaodong Yang; Ji Wang; Yuanpei Chen; |
| 444 | EVALUATION CARDS: An Interpretive Layer for AI Evaluation Reporting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present EVALUATION CARDS, an operational reporting layer that composes benchmark metadata, evaluation run data, and model metadata into a unified record. |
Avijit Ghosh; Anka Reuel-Lamparth; Wm. Matthew Kennedy; Jenny Chim; Yilin Huang; Srishti Yadav; Anastassia Kornilova; Damian Stachura; Kevin Klyman; Jennifer Mickel; Felix Friedrich; Max Lamparth; Jan Batzner; Anoop Mishra; Jeba Sania; Eliya Habba; Nathan Heath; Shalaleh Rismani; Usman Gohar; Andrea Loehr; David Manheim; Ruchira Dhar; Sree Harsha Nelaturu; Aarush Sinha; Leshem Choshen; Yixiong Hao; Yanan Long; Andrew Tran; Drishti Sharma; Ishan Khire; Amit Saha; Subramanyam Sahoo; Michael Hardy; Michael A. Riegler; Kabir Manghnani; Michelle Lin; Nancy Jiang; Asaf Yehudai; Nuno Moniz; Yacine Jernite; Zeerak Talat; Stella Biderman; Mubashara Akhtar; Jessica Ji; Aris Hofmann; Mykel J Kochenderfer; Sanmi Koyejo; Irene Solaiman; |
| 445 | Feature Recovery for Object Understanding Under Physical Transformation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Unlike standard image corruptions, fire damage changes the physical structure of the object itself. To study this setting, we introduce TRACE, a transformation-aware benchmark for post-fire object understanding. |
Aditi Tiwari; Sofia Stoica; Savya Khosla; David Forsyth; Heng Ji; |
| 446 | Endowing Your Vision-Language-Action Model with A Predictive Mind Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose PredMind, a plug-and-play predictive representation learning framework that treats future prediction as a training signal for the VLA backbone itself. |
Pengxiang Ding; Haoying Wang; Minghui Lin; Qishen Wang; Zhenyu Ding; Runze Suo; Xuanxuan An; Wenxuan Song; Fuhao Li; Han Zhao; Donglin Wang; Ning DING; |
| 447 | Activation Functions Shape Token Synchronization in Stochastic Transformer Dynamics Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study how the MLP activation function shapes the long-time geometry of token representations in a deep transformer. |
Nikita Karagodin; |
| 448 | Evaluating Depth and Breadth in Test-Time Scaling for Compositional Visual Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Test-time scaling is increasingly used to improve visual generation, yet it remains unclear how different scaling strategies should be evaluated when they trade off accuracy, latency, throughput, compute cost, and visual quality in different ways. We study this question through a systematic comparison of prominent test-time strategies including sampler-depth scaling, breadth-oriented parallel sampling, step-by-step construction, iterative refinement, and hybrid breadth–depth for compositional text-to-image generation. |
Shantanu Jaiswal; Mihir Prabhudesai; Nikash Bhardwaj; Zheyang Qin; Amir Zadeh; Chuan Li; Katerina Fragkiadaki; Deepak Pathak; |
| 449 | Spike-to-Field Mechanisms of Turbulence-Like Dynamics in Spatial Spiking Neural Networks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We build a spatial leaky integrate-and-fire spiking neural network and explicitly transform spikes into activity-density and Hilbert phase fields. |
Zequan Liu; |
| 450 | Firefly: Illuminating Verified Real-world Tool Call Data Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce \projectname, a large-scale synthetic and verifiable dataset for tool-calling agents built on real-world MCP servers.To support scalable offline training and evaluation, we further construct a retrieval-augmented simulator from explored trajectories, enabling reproducible tool execution without relying on live MCP servers. |
Yuxuan Lu; Ziyi Wang; Yingzhou Lu; Yisi Sang; Jiri Gesi; Xianfeng Tang; Yimeng Zhang; Zhenwei Dai; Hui Liu; Hanqing Lu; Chen Luo; Qi He; Benoit Dumoulin; Jing Huang; Dakuo Wang; |
| 451 | Reinforcement Learning for Code Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We make execution time learnable through three stages: (1) how code is tested, by building DMC-Optim with large optimization tests and a calibrated sandbox; (2) how speed is turned into reward, by composing correctness and speed in the RL environment and using an offline simulator to predict the most promising configurations; and (3) how the model learns from that reward, by adapting GRPO and evaluation to the sparser, noisier timed-execution setting. |
Pierre Chambon; Kunhao Zheng; Juliette Decugis; Benoît Sagot; Gabriel Synnaeve; |
| 452 | PIS: Pose-Interpolation Smoothness for Skinning Weight Refinement Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose \emph{Pose-Interpolation Smoothness} (PIS), a self-supervised, training-free, test-time refiner: minimize the per-vertex gap between the deformation at the SO(3) midpoint of two random poses and the linear average of the two endpoint deformations. |
Shaofeng Yin; Ting Lei; Yang Liu; |
| 453 | Thinking with Imagination: Agentic Visual Spatial Reasoning with World Simulators Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: They often struggle to infer unobserved layouts, maintain cross-view consistency, and reason from alternative viewpoints when only limited egocentric observations are available. In this work, we study this problem as thinking with imagination, where a VLM actively acquires imagined visual evidence by interacting with a world simulator during reasoning. |
Chenming Zhu; Jingli Lin; Yilin Long; Peizhou Cao; Tai WANG; Jiangmiao Pang; Xihui Liu; |
| 454 | When and How Much to Imagine: Adaptive Test-Time Scaling with World Models for Visual Spatial Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we present an in-depth analysis of test-time visual imagination as a controllable resource for spatial reasoning. |
Shoubin Yu; Yue Zhang; Zun Wang; Jaehong Yoon; Huaxiu Yao; Mingyu Ding; Mohit Bansal; |
| 455 | PolyTopoBench: A Benchmark for Complex Vector Polygon Generation from Remote Sensing Imagery Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose PolyTopoBench, a unified evaluation framework for vector polygon generation from RS images with explicit emphasis on complex polygons. |
Zeping Liu; Ni Lao; Weiwei Sun; Gil Wolff; Yiqun Xie; Liang Zhao; Junfeng Jiao; Gengchen Mai; |
| 456 | STILL: Selecting Tokens for Intra-Layer Hybrid Attention to Linearize LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Meanwhile, linear attention further suffers from distribution shift caused by learnable feature maps that distort pretrained feature magnitudes. Motivated by these limitations, we propose STILL, an intra-layer hybrid linearization framework for efficiently linearizing LLMs. |
Weikang Meng; Liangyu Huo; Yadan Luo; Jiawen Guan; Jingyi Zhang; Yingjian Li; Zheng Zhang; |
| 457 | Structural Blindness in Latent Data Assimilation: Representation Geometry Misleads Sensor Design Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study latent Kalman-type data assimilation, where an encoder–decoder pair is trained without knowledge of the sensor used to collect observations. |
Yongquan Qu; Pierre Gentine; |
| 458 | Incorporating Hierarchical Semantics in Sparse Autoencoder Architectures Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we introduce a modified SAE architecture that explicitly models a semantic hierarchy of concepts. |
Mark Muchane; Sean M Richardson; Kiho Park; Victor Veitch; |
| 459 | GRNAgent: A Multimodal Graph Reasoning Agent for Gene Regulatory Network Inference Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose GRNAgent, an automated multi-agent framework for multi-omic GRN inference that coordinates evidence acquisition, quality control, modality integration, model inference, benchmarking, and post hoc validation. |
Akshata Hegde; Kyler Zook; Yanli Wang; Jianlin Cheng; |
| 460 | A Mechanistic Investigation of Theory of Mind in A Large Language Model Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Large language models successfully solve classic theory-of-mind tasks adapted from cognitive science, but the mechanisms underlying this ability remain unclear. |
Idil K Sahin; Steven M Frankland; Taylor Webb; |
| 461 | Binding Visual Features Point By Point Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Here, we investigate this question. We find that learning to point-via-text induces an internal visual search routine, and we characterize the mechanisms that support this procedure. |
Udith Haputhanthri; Declan Campbell; Rim Assouel; Jonathan D Cohen; Taylor Webb; |
| 462 | DexMirror: Real-to-Sim Scene Mirroring for Sim-to-Real Dexterous Manipulation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present **DexMirror**, a unified real-to-sim-to-real framework that enables photorealistic simulation learning and reliable deployment for dexterous manipulation. |
Pu Hua; Zhecheng Yuan; Meng Chu; Peiqi Duan; Huazhe Xu; |
| 463 | Test-Time Personalization: A Diagnostic Framework and Probabilistic Fix for Scaling Failures Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we study \textbf{Test-Time Personalization (TTP)} along an unexplored axis: scaling inference-time computation by sampling $N$ candidates from a personalized policy model and selecting the best with a personalized reward model. |
Linhai Zhang; Yulan He; |
| 464 | Unifying Sparsity and Discreteness: One-Shot Pruning for Quantized LLMs Via Discrete Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: The discreteness of quantization shatters their required variable continuity, inevitably collapsing this integration into a decoupled pruning-quantization pipeline with inherently suboptimal outcomes. To tackle this problem, we propose Quantization-aware One-shot Pruning (QOP), which directly optimizes the quantized weights under sparsity and discreteness constraints, thereby explicitly capturing the impact of quantization on pruning within its objective. |
Haozhen Zhang; Hanyuan Zheng; Teng Hou; Zhaogeng Liu; Yi Chang; Bin Gu; |
| 465 | Reasoning Via Test-Time Instance-Level Policy Gradient in Latent Space Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we introduce LatentSeek, a novel framework designed to enhance the reasoning capabilities of LLMs through Test-Time Instance-level Policy Gradient within the model’s latent space—thus complementing explicit natural-language steps. |
Hengli Li; Chenxi Li; Tong Wu; Xuekai Zhu; Yuxuan Wang; Zhaoxin Yu; Eric Hanchen Jiang; Song-Chun Zhu; Zixia Jia; Ying Nian Wu; Zilong Zheng; |
| 466 | Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While Twisted Sequential Monte Carlo (SMC) offers asymptotic exactness for this task, estimating the optimal twist function in discrete state spaces necessitates costly Monte Carlo approximations, resulting a severe computational bottleneck at inference. To overcome this limitation, we introduce Contrastive Distribution Matching (CDM), a novel framework that amortizes the cost of SMC inference by learning a parameterized twist function via positive and negative samples. |
Jaihoon Kim; Taehoon Yoon; Prin Phunyaphibarn; Seungjun Kim; Morteza Mardani; Minhyuk Sung; |
| 467 | Sparse Reward Subsystem in Large Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we show that reward-related information is concentrated in a sparse subset of neurons. |
Guowei Xu; Mert Yuksekgonul; James Zou; |
| 468 | MonarchRT: Efficient Attention for Real-Time Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Specifically, we observe that video attention is not reliably sparse, but instead combines pronounced periodic structure driven by spatiotemporal position with dynamic, sparse semantic correspondences and dense mixing, exceeding the representational capacity of even oracle top-k attention. Building on this insight, we propose MonarchRT, a structured attention parameterization for video diffusion models that factorizes attention using Monarch matrices. |
Krish Agarwal; Zhuoming Chen; cheng Luo; Yongqi Chen; Haizhong Zheng; Xun Huang; Atri Rudra; Beidi Chen; |
| 469 | The AI Observatory: A Public Measure of Real-World AI Use Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce the AI Observatory, a public measurement platform that aggregates the most comprehensive set of real AI conversation sources, and develop a common taxonomy of 145 features, spanning function, topic, sensitive use, interaction style, multi-turn dynamics, and conversation structure. |
Shayne Longpre; Anka Reuel-Lamparth; Dayeon Ki; Zhiping Zhang; Cathy Fang; Chuanyang Jin; Jennifer Mickel; Cedric Whitney; Niloofar Mireshghallah; Victor Ojewale; Megan Richards; Ariel N. Lee; Alex Pentland; Tianshi Li; Yuntian Deng; Sara Hooker; Mykel J Kochenderfer; Sanmi Koyejo; |
| 470 | Learning to Solve Compositional Geometry Routing Problems Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Beyond standard point-based routing, CGRP with non-point tasks can be inherently asymmetric, tightly coupled travel routes with the intrinsic path, and enlarges the action space with numerous feasible yet often irrelevant options, thereby posing significant challenges for both representation learning and decision-making. To address these challenges, we propose DiCon, a differential attention–assisted solver with contrastive learning, as a plug-and-play framework that tackles the problem from two complementary angles. |
Mingfeng Fan; Jianan Zhou; Jiaqi Cheng; Yifeng Zhang; Jie Zhang; Guillaume Sartoretti; |
| 471 | SWE-Protégé: Learning to Selectively Collaborate With An Expert Unlocks Small Language Models As Software Engineering Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce SWE-Protégé, a post-training framework that reframes software repair as an expert–protégé collaboration problem. |
Patrick Tser Jern Kon; Archana Pradeep; Ang Chen; Alex Ellis; Warren Hunt; Zijian Wang; John Yang; Samuel Thompson; |
| 472 | VeriVul: A Verification-Guided Framework for Generating Realistic Vulnerability Benchmarks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present VeriVul, a framework that produces formally verified vulnerability data on demand.We release the framework, the slicer, the verification harness, and the generated dataset to support contamination-resilient evaluation of code LLMs as vulnerability analysts. |
Ahmed Lekssays; Hamza Mouhcine; Issa Khalil; |
| 473 | OpenWebRL: Demystifying Online Multi-turn Reinforcement Learning for Visual Web Agents Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we introduce OpenWebRL, an open framework for training visual web agents with online multi-turn RL on real websites.We will release our training data, models, and code to support future research. |
Rui Yang; Qianhui Wu; Yuxi Chen; Hao Bai; Wenlin Yao; Hao Cheng; Baolin Peng; Huan Zhang; Tong Zhang; Jianfeng Gao; |
| 474 | LLM-Based Multi-Agent Blackboard System for Information Discovery in Data Science Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a novel multi-agent paradigm inspired by the blackboard architecture for traditional AI models. |
Alireza Salemi; Mihir Parmar; Palash Goyal; Yiwen Song; Jinsung Yoon; Hamed Zamani; Tomas Pfister; Hamid Palangi; |
| 475 | Stitched Value Model for Diffusion Alignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Here, we propose StitchVM (Stitched Value Model), a model stitching framework that efficiently transfers reward models pretrained for clean images to the noisy latent regime. |
Hyojun Go; Hyungjin Chung; Prune Truong; Goutam Bhat; Li Mi; Zhaochong An; Zixiang Zhao; Dominik Narnhofer; Serge Belongie; Federico Tombari; Konrad Schindler; |
| 476 | Instrumental Choices: Measuring The Propensity of LLM Agents to Pursue Instrumental Behaviors Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce a benchmark for measuring model propensity for instrumental convergence (IC) behaviour in terminal-based agents. |
Jonas Wiedermann-Möller; Leonard Dung; Maksym Andriushchenko; |
| 477 | Read It Back: Pretrained MLLMs Are Zero-shot Reward Models for Text-to-Image Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose **SpectraReward**, a training-free reward function that turns pretrained MLLMs into off-the-shelf reward models for image-generation reinforcement learning. |
Runhui Huang; Qihui Zhang; Zhe Liu; Yu Gao; Jie Wu; Hengshuang Zhao; |
| 478 | Vortex: Efficient and Programmable Sparse Attention Serving Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, deploying and evaluating new sparse attention algorithms at scale remains highly engineering-intensive, slowing both human researchers and AI agents in exploring the sparse attention design. To address this challenge, we present Vortex, a system that combines a Python-embedded frontend language atop a page-centric tensor abstraction for expressing a broad range of sparse attention algorithms, with an efficient backend tightly integrated into modern LLM serving stacks. |
Zhuoming Chen; Xinrui Zhong; Qilong Feng; Ranajoy Sadhukhan; Yang Zhou; Michael Shieh; Zhihao Jia; Beidi Chen; |
| 479 | PhysFormer: Learning to Simulate Mechanics in World Space Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present PhysFormer, a physics-grounded diffusion transformer for generating 4D multi-object mesh dynamics directly in world coordinates. |
Yiming Chen; Yushi LAN; Andrea Vedaldi; |
| 480 | ViroGym: Realistic Large-Scale Benchmarks for Evaluating Viral Proteins Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Here we introduce ViroGym, a comprehensive benchmark evaluating pLMs across three tasks: 79 deep mutational scanning (DMS) assays covering eukaryotic viruses with 552,065 mutated sequences across 7 phenotypic readouts, 21 influenza neutralisation tasks, and a real-world pandemic prediction task for SARS-CoV-2. |
Yichen Zhou; Jonathan L Golob; Amir Mohammad Karimi Mamaghan; Stefan Bauer; Patrick Schwab; |
| 481 | JRDB-AVR: An Active Visual Reasoning Benchmark for Real-World Embodied Environments Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Current visual reasoning benchmarks largely evaluate passive observations and final answers, overlooking settings that require active perception and evidence acquisition. We introduce JRDB-AVR, a benchmark built from real-world human-scene robotics data that turns this gap into an explicit evaluation: a system receives a visual reasoning question, requests bounded observations by timestamp and viewing angle, and is evaluated on both the final answer and the grounded visual evidence supporting it. |
Zhixi Cai; Fucai Ke; Sukai Huang; Maria Garcia de la Banda; Peter Stuckey; Reza Haffari; Hamid Rezatofighi; |
| 482 | Towards High Semantic Fidelity: Hyperdimensional Symbolic Messages in Multi-Agent Communication Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Yet existing numerical and neural message formats cannot simultaneously achieve both. To address this, we propose a novel message format-\textbf{H}yperdimensional \textbf{S}ymbolic \textbf{M}essages (HSM), and its corresponding communication framework \textbf{Herm}. |
Shoucheng Song; Youfang Lin; Chang Yao; Hao Wu; Sheng Han; Kai Lv; |
| 483 | Boosting Multiagent Reinforcement Learning at High Replay Ratios with Ensemble Reset Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we show that high replay ratios induce a severe dormant neuron problem in the centralized global Q-network of MARL, where a large fraction of neurons become inactive, thereby reducing network capacity and destabilizing learning. |
Yaodong Yang; Hongyao Tang; Guangyong Chen; Pheng-Ann Heng; |
| 484 | When Are Teacher Tokens Reliable? Position-Weighted On-Policy Self-Distillation for Reasoning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing entropy-based OPD methods relax this uniformity by modulating token-level supervision with teacher entropy, but high teacher entropy in reasoning has an ambiguous reliability meaning: it can reflect either non-viable uncertainty or benign solution diversity. To identify this phenomenon, we introduce a branch-viability diagnostic. |
Xiaogeng Liu; Xinyan Wang; Yingzi Ma; Yechao Zhang; Chaowei Xiao; |
| 485 | How Fine-Tuning Objectives Shape Layer-Wise Information in LLM Hallucination Detection Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: By truncating roughly 31\% of the network, we obtain early-exit models that match or exceed full-model performance on Qwen3-4B and 8B. |
Kaiyang Wan; Forrest Bao; Amin Ahmad; Yuxia Wang; |
| 486 | Uncertainty Quantification for Large Language Diffusion Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we present the first systematic study of UQ for LLDMs and propose lightweight, zero-shot uncertainty signals derived from the iterative denoising process, leveraging intermediate generations, token remasking dynamics, and denoising complexity. |
Artem Vazhentsev; Vladislav Smirnov; David Li; Maxim Panov; Timothy Baldwin; Artem Shelmanov; |
| 487 | Agentic-imodels: Evolving Agentic Interpretability Tools Via Autoresearch Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, current ADS systems use statistical tools designed to be interpretable by humans, rather than interpretable by agents. To address this, we introduce Agentic-imodels, an agentic autoresearch loop that evolves data-science tools designed to be interpretable by agents. |
Chandan Singh; Yan Shuo Tan; Weijia Xu; Zelalem Gero; Weiwei Yang; Michel Galley; Jianfeng Gao; |
| 488 | Rethinking Projector Training in Multimodal LLMs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: These results suggest that projector training acts primarily as a coarse alignment within a large solution space. Motivated by these findings, we propose PORTAL, a training-free projector initialization that requires neither paired image-caption data nor gradient-based optimization, but only per-modality summary statistics. |
Fabian Gröger; Naël Ouerghemi; Shuo Wen; Maria Brbic; |
| 489 | To Align or Not To Align: Check Your COMPASS Before You Train Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose COMPASS (COMPatibility ASsessment Score), a theoretically grounded compatibility diagnostic that estimates whether a model pair will be compatible prior to training or merging. |
Fabian Gröger; Petar Damjanović; Maria Brbic; |
| 490 | Scaffold3D: SfM-Conditioned Pointmap Prediction for Multi-View 3D Reconstruction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Structure-from-motion (SfM) offers a complementary geometric signal through camera poses and sparse 3D points estimated by correspondence filtering and optimization. To propagate this globally consistent \emph{SfM scaffold} into dense geometry, we introduce \textbf{Scaffold3D}, an SfM-conditioned reconstruction framework that injects image patch-aligned SfM point tokens into a pairwise feed-forward pointmap predictor. |
Frano Rajič; Yutong Chen; Haofei Xu; Zador Pataki; Marc Pollefeys; Siyu Tang; |
| 491 | Beyond Pairwise Supervision: Spectral Characteristic Matching for Data-Efficient Multimodal Alignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we propose $\textbf{Spec}$tral Characteristic Distribution $\textbf{Align}$ment ($\textbf{SpecAlign}$), a lightweight spectral alignment framework that reduces cross-modal mismatch by comparing image and text characteristic functions over learnable spectral queries. |
Xinchao Wang; Shuang Li; Wei Chen; Jingxuan Kang; Fuzhen Zhuang; deqing wang; |
| 492 | Counterfactual Instruction Grounding for Vision-Language-Action Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We address this by recasting instruction grounding as identifying the commanded instruction among counterfactual alternatives. Based on this view, we propose Counterfactual Instruction Grounding (CIG), a contrastive objective that encourages the generated trajectory to remain identifiable with the commanded instruction among counterfactual alternatives. |
Shiyu Liu; Yu Zhou; Meng Liu; Xuanming Guo; Liqiang Nie; |
| 493 | Efficient Benchmarking Is Just Feature Selection and Multiple Regression Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Additionally, using an information-theoretic feature-selection algorithm called minimum redundancy maximum relevance (mRMR), we can further improve upon these methods by selecting question subsets that will be maximally useful for prediction. Except in very data-poor settings, these approaches consistently achieve smaller prediction errors (in both MAE and RMSE), and greater ranking correlation between predicted and true scores (in both Spearman $\rho$ and Kendall $\tau$) across a range of benchmarks using both binary and continuous metrics. |
Sam Bowyer; Acyr Locatelli; Kris Cao; |
| 494 | DSAQuant: Denoising-Stage-Aligned Quantization-Aware Training for Video Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: % However, existing QAT methods suffer from a distinctive challenge in VDMs: while they often preserve prompt semantics, global layout, and coarse motion, the quantized model severely degrades visual details, texture fidelity, and sharpness. In this paper, we trace this degradation to the timestep-agnostic design of conventional quantization pipelines, which overlooks the stage-wise functionality of video denoising. |
Shuaiting Li; Zelin Gao; Haibin Shen; Yujun Shen; Haotong Qin; Yinghao Xu; |
| 495 | AI Control for Sandbagging on Fuzzy Tasks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: To understand sandbagging on fuzzy tasks, we introduce a novel AI control framework that considers sandbagging as an adversarial game between a blue team and a red team. |
Mikhail Terekhov; Caglar Gulcehre; Vivek Hebbar; Joe Benton; |
| 496 | AffectGPT-RL: Revealing Roles of Reinforcement Learning in Open-Vocabulary Emotion Recognition Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Unlike traditional discriminative methods, OV-MER leverages generative models to capture the full spectrum of emotions and employs emotion wheels (EWs) for metric calculation. |
Zheng Lian; Fan Zhang; Lan Chen; Yazhou Zhang; Rui Liu; Jinyang Wu; Haoyu Chen; Xiaobai Li; Xiaojiang Peng; Bin He; Jianhua Tao; |
| 497 | RefDecoder: Enhancing Visual Generation with Conditional Video Decoding Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce RefDecoder, a reference-conditioned video VAE decoder by injecting the reference image directly into the decoding process via reference attention. |
Xiang Fan; Yuheng Wang; Bohan Fang; Jason Ren; Ranjay Krishna; |
| 498 | State Evolution Awareness for Category-agnostic 3D Point Cloud Tracking Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing 3D point cloud tracking methods have not fully exploited temporal information in point cloud sequences, making it difficult to effectively capture the continuous evolution of target states and thereby limiting tracking stability in scenarios with large cross-category geometric variation. To address this issue, we propose TETrack3D, a framework for category-agnostic 3D point cloud tracking. |
Xiantao Hu; Ying Tai; Jinxia Xie; Bineng Zhong; Guangwei Gao; Jianjun Qian; Jian Yang; |
| 499 | VFIG: Vectorizing Complex Figures in SVG with Vision-Language Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We close both gaps with paired contributions targeting diagram-centric figures (e.g., model architectures, flowcharts, schematics). For training, we introduce VFIG-Data, the largest figure-to-SVG dataset of its kind at 66K pairs, combining real paper figures converted via a describe-and-generate pipeline with programmatic diagrams that supply noise-free supervision over arrow styles, fonts, and geometry. |
Qijia He; Xunmei Liu; Hammaad Memon; Ziang Li; Zixian Ma; Jaemin Cho; Jason Ren; Daniel Weld; Ranjay Krishna; |
| 500 | StaDy: Factorizing The World Into Static and Dynamic Via Likelihood Matching Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing approaches address this issue through restrictive bottlenecks, which reduce leakage at the cost of limiting expressivity. We propose StaDy, a regularization framework that instead enforces conditional informativeness: the latent code should aid prediction only when paired with the source state, while remaining uninformative about the target on its own. |
Thomas Ressler-Antal; Frank Fundel; Malek Ben Alaya; Stefan Andreas Baumann; Björn Ommer; |
This table only includes 500 papers selected by our daily digest algorithm. To continue with the full list (~9,000 papers), please visit Paper Digest: NeurIPS-2026 (Full List).