Paper Digest: NeurIPS 2020 Highlights
Download NIPS-2020-Paper-Digests.pdf– Highlights of all 1,899 NeurIPS-2020 papers. Readers can choose to read all these highlights on our console as well, which allows users to filter out papers using keywords and find related papers, patents, etc. In addition, we identified a large number of papers that have published their code and data. Here is the (code/data download page).
The Conference on Neural Information Processing Systems (NIPS) is one of the top machine learning conferences in the world. In 2020, it is to be held online due to coivd-19 pandemic.
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 / summaries to quickly get the main idea of each paper.
If you do not want to miss any interesting academic paper, you are welcome to sign up our free daily paper digest service to get updates on new papers published in your area every day. You are also welcome to follow us on Twitter and Linkedin to get updated with new conference digests.
Paper Digest Team
team@paperdigest.org
TABLE 1: Paper Digest: NeurIPS 2020 Highlights
| Paper | Author(s) | Code | |
|---|---|---|---|
| 1 | A Graph Similarity For Deep Learning Highlight: We adopt kernel distance and propose transform-sum-cat as an alternative to aggregate-transform to reflect the continuous similarity between the node neighborhoods in the neighborhood aggregation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Seongmin Ok; | |
| 2 | An Unsupervised Information-Theoretic Perceptual Quality Metric Highlight: We combine recent advances in information-theoretic objective functions with a computational architecture informed by the physiology of the human visual system and unsupervised training on pairs of video frames, yielding our Perceptual Information Metric (PIM). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sangnie Bhardwaj; Ian Fischer; Johannes Ball�; Troy Chinen; | |
| 3 | Self-Supervised MultiModal Versatile Networks Highlight: In this work, we learn representations using self-supervision by leveraging three modalities naturally present in videos: visual, audio and language streams. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jean-Baptiste Alayrac; Adria Recasens; Rosalia Schneider; Relja Arandjelovic; Jason Ramapuram; Jeffrey De Fauw; Lucas Smaira; Sander Dieleman; Andrew Zisserman; | |
| 4 | Benchmarking Deep Inverse Models Over Time, And The Neural-Adjoint Method Highlight: We consider the task of solving generic inverse problems, where one wishes to determine the hidden parameters of a natural system that will give rise to a particular set of measurements. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Simiao Ren; Willie Padilla; Jordan Malof; | |
| 5 | Off-Policy Evaluation And Learning For External Validity Under A Covariate Shift Highlight: In this paper, we derive the efficiency bound of OPE under a covariate shift. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Masatoshi Uehara; Masahiro Kato; Shota Yasui; | |
| 6 | Neural Methods For Point-wise Dependency Estimation Highlight: In this work, instead of estimating the expected dependency, we focus on estimating point-wise dependency (PD), which quantitatively measures how likely two outcomes co-occur. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yao-Hung Hubert Tsai; Han Zhao; Makoto Yamada; Louis-Philippe Morency; Russ R. Salakhutdinov; | |
| 7 | Fast And Flexible Temporal Point Processes With Triangular Maps Highlight: By exploiting the recent developments in the field of normalizing flows, we design TriTPP – a new class of non-recurrent TPP models, where both sampling and likelihood computation can be done in parallel. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Oleksandr Shchur; Nicholas Gao; Marin Bilo�; Stephan G�nnemann; | |
| 8 | Backpropagating Linearly Improves Transferability Of Adversarial Examples Highlight: In this paper, we study the transferability of such examples, which lays the foundation of many black-box attacks on DNNs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yiwen Guo; Qizhang Li; Hao Chen; | code |
| 9 | PyGlove: Symbolic Programming For Automated Machine Learning Highlight: In this paper, we introduce a new way of programming AutoML based on symbolic programming. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Daiyi Peng; Xuanyi Dong; Esteban Real; Mingxing Tan; Yifeng Lu; Gabriel Bender; Hanxiao Liu; Adam Kraft; Chen Liang; Quoc Le; | |
| 10 | Fourier Sparse Leverage Scores And Approximate Kernel Learning Highlight: We prove new explicit upper bounds on the leverage scores of Fourier sparse functions under both the Gaussian and Laplace measures. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tamas Erdelyi; Cameron Musco; Christopher Musco; | |
| 11 | Improved Algorithms For Online Submodular Maximization Via First-order Regret Bounds Highlight: In this work, we give a general approach for improving regret bounds in online submodular maximization by exploiting “first-order” regret bounds for online linear optimization. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Nicholas Harvey; Christopher Liaw; Tasuku Soma; | |
| 12 | Synbols: Probing Learning Algorithms With Synthetic Datasets Highlight: In this sense, we introduce Synbols — Synthetic Symbols — a tool for rapidly generating new datasets with a rich composition of latent features rendered in low resolution images. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Alexandre Lacoste; Pau Rodr�guez L�pez; Frederic Branchaud-Charron; Parmida Atighehchian; Massimo Caccia; Issam Hadj Laradji; Alexandre Drouin; Matthew Craddock; Laurent Charlin; David V�zquez; | |
| 13 | Adversarially Robust Streaming Algorithms Via Differential Privacy Highlight: We establish a connection between adversarial robustness of streaming algorithms and the notion of differential privacy. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Avinatan Hasidim; Haim Kaplan; Yishay Mansour; Yossi Matias; Uri Stemmer; | |
| 14 | Trading Personalization For Accuracy: Data Debugging In Collaborative Filtering Highlight: In this paper, we propose a data debugging framework to identify overly personalized ratings whose existence degrades the performance of a given collaborative filtering model. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Long Chen; Yuan Yao; Feng Xu; Miao Xu; Hanghang Tong; | |
| 15 | Cascaded Text Generation With Markov Transformers Highlight: This work proposes an autoregressive model with sub-linear parallel time generation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yuntian Deng; Alexander Rush; | |
| 16 | Improving Local Identifiability In Probabilistic Box Embeddings Highlight: In this work we model the box parameters with min and max Gumbel distributions, which were chosen such that the space is still closed under the operation of intersection. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shib Dasgupta; Michael Boratko; Dongxu Zhang; Luke Vilnis; Xiang Li; Andrew McCallum; | |
| 17 | Permute-and-Flip: A New Mechanism For Differentially Private Selection Highlight: In this work, we propose a new mechanism for this task based on a careful analysis of the privacy constraints. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ryan McKenna; Daniel R. Sheldon; | |
| 18 | Deep Reconstruction Of Strange Attractors From Time Series Highlight: Inspired by classical analysis techniques for partial observations of chaotic attractors, we introduce a general embedding technique for univariate and multivariate time series, consisting of an autoencoder trained with a novel latent-space loss function. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
William Gilpin; | |
| 19 | Reciprocal Adversarial Learning Via Characteristic Functions Highlight: We generalise this by comparing the distributions rather than their moments via a powerful tool, i.e., the characteristic function (CF), which uniquely and universally comprising all the information about a distribution. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shengxi Li; Zeyang Yu; Min Xiang; Danilo Mandic; | |
| 20 | Statistical Guarantees Of Distributed Nearest Neighbor Classification Highlight: Through majority voting, the distributed nearest neighbor classifier achieves the same rate of convergence as its oracle version in terms of the regret, up to a multiplicative constant that depends solely on the data dimension. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jiexin Duan; Xingye Qiao; Guang Cheng; | |
| 21 | Stein Self-Repulsive Dynamics: Benefits From Past Samples Highlight: We propose a new Stein self-repulsive dynamics for obtaining diversified samples from intractable un-normalized distributions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Mao Ye; Tongzheng Ren; Qiang Liu; | |
| 22 | The Statistical Complexity Of Early-Stopped Mirror Descent Highlight: In this paper, we study the statistical guarantees on the excess risk achieved by early-stopped unconstrained mirror descent algorithms applied to the unregularized empirical risk with the squared loss for linear models and kernel methods. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tomas Vaskevicius; Varun Kanade; Patrick Rebeschini; | |
| 23 | Algorithmic Recourse Under Imperfect Causal Knowledge: A Probabilistic Approach Highlight: To address this limitation, we propose two probabilistic approaches to select optimal actions that achieve recourse with high probability given limited causal knowledge (e.g., only the causal graph). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Amir-Hossein Karimi; Julius von K�gelgen; Bernhard Sch�lkopf; Isabel Valera; | |
| 24 | Quantitative Propagation Of Chaos For SGD In Wide Neural Networks Highlight: In this paper, we investigate the limiting behavior of a continuous-time counterpart of the Stochastic Gradient Descent (SGD) algorithm applied to two-layer overparameterized neural networks, as the number or neurons (i.e., the size of the hidden layer) $N \to \plusinfty$. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Valentin De Bortoli; Alain Durmus; Xavier Fontaine; Umut Simsekli; | |
| 25 | A Causal View On Robustness Of Neural Networks Highlight: We present a causal view on the robustness of neural networks against input manipulations, which applies not only to traditional classification tasks but also to general measurement data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Cheng Zhang; Kun Zhang; Yingzhen Li; | |
| 26 | Minimax Classification With 0-1 Loss And Performance Guarantees Highlight: This paper presents minimax risk classifiers (MRCs) that do not rely on a choice of surrogate loss and family of rules. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Santiago Mazuelas; Andrea Zanoni; Aritz P�rez; | |
| 27 | How To Learn A Useful Critic? Model-based Action-Gradient-Estimator Policy Optimization Highlight: In this paper, we propose MAGE, a model-based actor-critic algorithm, grounded in the theory of policy gradients, which explicitly learns the action-value gradient. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Pierluca D'Oro; Wojciech Jaskowski; | |
| 28 | Coresets For Regressions With Panel Data Highlight: This paper introduces the problem of coresets for regression problems to panel data settings. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lingxiao Huang; K Sudhir; Nisheeth Vishnoi; | |
| 29 | Learning Composable Energy Surrogates For PDE Order Reduction Highlight: To address this, we leverage parametric modular structure to learn component-level surrogates, enabling cheaper high-fidelity simulation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Alex Beatson; Jordan Ash; Geoffrey Roeder; Tianju Xue; Ryan P. Adams; | |
| 30 | Efficient Contextual Bandits With Continuous Actions Highlight: We create a computationally tractable learning algorithm for contextual bandits with continuous actions having unknown structure. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Maryam Majzoubi; Chicheng Zhang; Rajan Chari; Akshay Krishnamurthy; John Langford; Aleksandrs Slivkins; | |
| 31 | Achieving Equalized Odds By Resampling Sensitive Attributes Highlight: We present a flexible framework for learning predictive models that approximately satisfy the equalized odds notion of fairness. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yaniv Romano; Stephen Bates; Emmanuel Candes; | |
| 32 | Multi-Robot Collision Avoidance Under Uncertainty With Probabilistic Safety Barrier Certificates Highlight: This paper aims to propose a collision avoidance method that accounts for both measurement uncertainty and motion uncertainty. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Wenhao Luo; Wen Sun; Ashish Kapoor; | |
| 33 | Hard Shape-Constrained Kernel Machines Highlight: In this paper, we prove that hard affine shape constraints on function derivatives can be encoded in kernel machines which represent one of the most flexible and powerful tools in machine learning and statistics. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Pierre-Cyril Aubin-Frankowski; Zoltan Szabo; | |
| 34 | A Closer Look At The Training Strategy For Modern Meta-Learning Highlight: The support/query (S/Q) episodic training strategy has been widely used in modern meta-learning algorithms and is believed to improve their generalization ability to test environments. This paper conducts a theoretical investigation of this training strategy on generalization. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
JIAXIN CHEN; Xiao-Ming Wu; Yanke Li; Qimai LI; Li-Ming Zhan; Fu-lai Chung; | |
| 35 | On The Value Of Out-of-Distribution Testing: An Example Of Goodhart's Law Highlight: We provide short- and long-term solutions to avoid these pitfalls and realize the benefits of OOD evaluation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Damien Teney; Ehsan Abbasnejad; Kushal Kafle; Robik Shrestha; Christopher Kanan; Anton van den Hengel; | |
| 36 | Generalised Bayesian Filtering Via Sequential Monte Carlo Highlight: We introduce a framework for inference in general state-space hidden Markov models (HMMs) under likelihood misspecification. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ayman Boustati; Omer Deniz Akyildiz; Theodoros Damoulas; Adam Johansen; | |
| 37 | Deterministic Approximation For Submodular Maximization Over A Matroid In Nearly Linear Time Highlight: We study the problem of maximizing a non-monotone, non-negative submodular function subject to a matroid constraint. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kai Han; zongmai Cao; Shuang Cui; Benwei Wu; | |
| 38 | Flows For Simultaneous Manifold Learning And Density Estimation Highlight: We introduce manifold-learning flows (?-flows), a new class of generative models that simultaneously learn the data manifold as well as a tractable probability density on that manifold. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Johann Brehmer; Kyle Cranmer; | |
| 39 | Simultaneous Preference And Metric Learning From Paired Comparisons Highlight: In this paper, we consider the problem of learning an ideal point representation of a user’s preferences when the distance metric is an unknown Mahalanobis metric. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Austin Xu; Mark Davenport; | |
| 40 | Efficient Variational Inference For Sparse Deep Learning With Theoretical Guarantee Highlight: In this paper, we train sparse deep neural networks with a fully Bayesian treatment under spike-and-slab priors, and develop a set of computationally efficient variational inferences via continuous relaxation of Bernoulli distribution. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jincheng Bai; Qifan Song; Guang Cheng; | |
| 41 | Learning Manifold Implicitly Via Explicit Heat-Kernel Learning Highlight: In this paper, we propose the concept of implicit manifold learning, where manifold information is implicitly obtained by learning the associated heat kernel. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yufan Zhou; Changyou Chen; Jinhui Xu; | |
| 42 | Deep Relational Topic Modeling Via Graph Poisson Gamma Belief Network Highlight: To better utilize the document network, we first propose graph Poisson factor analysis (GPFA) that constructs a probabilistic model for interconnected documents and also provides closed-form Gibbs sampling update equations, moving beyond sophisticated approximate assumptions of existing RTMs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chaojie Wang; Hao Zhang; Bo Chen; Dongsheng Wang; Zhengjue Wang; Mingyuan Zhou; | |
| 43 | One-bit Supervision For Image Classification Highlight: This paper presents one-bit supervision, a novel setting of learning from incomplete annotations, in the scenario of image classification. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
hengtong hu; Lingxi Xie; Zewei Du; Richang Hong; Qi Tian; | |
| 44 | What Is Being Transferred In Transfer Learning? Highlight: In this paper, we provide new tools and analysis to address these fundamental questions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Behnam Neyshabur; Hanie Sedghi; Chiyuan Zhang; | |
| 45 | Submodular Maximization Through Barrier Functions Highlight: In this paper, we introduce a novel technique for constrained submodular maximization, inspired by barrier functions in continuous optimization. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ashwinkumar Badanidiyuru; Amin Karbasi; Ehsan Kazemi; Jan Vondrak; | |
| 46 | Neural Networks With Recurrent Generative Feedback Highlight: The proposed framework, termed Convolutional Neural Networks with Feedback (CNN-F), introduces a generative feedback with latent variables to existing CNN architectures, where consistent predictions are made through alternating MAP inference under a Bayesian framework. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yujia Huang; James Gornet; Sihui Dai; Zhiding Yu; Tan Nguyen; Doris Tsao; Anima Anandkumar; | |
| 47 | Learning To Extrapolate Knowledge: Transductive Few-shot Out-of-Graph Link Prediction Highlight: Motivated by this challenge, we introduce a realistic problem of few-shot out-of-graph link prediction, where we not only predict the links between the seen and unseen nodes as in a conventional out-of-knowledge link prediction task but also between the unseen nodes, with only few edges per node. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jinheon Baek; Dong Bok Lee; Sung Ju Hwang; | |
| 48 | Exploiting Weakly Supervised Visual Patterns To Learn From Partial Annotations Highlight: Instead, in this paper, we exploit relationships among images and labels to derive more supervisory signal from the un-annotated labels. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kaustav Kundu; Joseph Tighe; | |
| 49 | Improving Inference For Neural Image Compression Highlight: We consider the problem of lossy image compression with deep latent variable models. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yibo Yang; Robert Bamler; Stephan Mandt; | |
| 50 | Neuron Merging: Compensating For Pruned Neurons Highlight: In this work, we propose a novel concept of neuron merging applicable to both fully connected layers and convolution layers, which compensates for the information loss due to the pruned neurons/filters. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Woojeong Kim; Suhyun Kim; Mincheol Park; Geunseok Jeon; | code |
| 51 | FixMatch: Simplifying Semi-Supervised Learning With Consistency And Confidence Highlight: In this paper we propose FixMatch, an algorithm that is a significant simplification of existing SSL methods. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kihyuk Sohn; David Berthelot; Nicholas Carlini; Zizhao Zhang; Han Zhang; Colin A. Raffel; Ekin Dogus Cubuk; Alexey Kurakin; Chun-Liang Li; | code |
| 52 | Reinforcement Learning With Combinatorial Actions: An Application To Vehicle Routing Highlight: We develop a framework for value-function-based deep reinforcement learning with a combinatorial action space, in which the action selection problem is explicitly formulated as a mixed-integer optimization problem. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Arthur Delarue; Ross Anderson; Christian Tjandraatmadja; | |
| 53 | Towards Playing Full MOBA Games With Deep Reinforcement Learning Highlight: In this paper, we propose a MOBA AI learning paradigm that methodologically enables playing full MOBA games with deep reinforcement learning. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Deheng Ye; Guibin Chen; Wen Zhang; Sheng Chen; Bo Yuan; Bo Liu; Jia Chen; Zhao Liu; Fuhao Qiu; Hongsheng Yu; Yinyuting Yin; Bei Shi; Liang Wang; Tengfei Shi; Qiang Fu; Wei Yang; Lanxiao Huang; Wei Liu; | |
| 54 | Rankmax: An Adaptive Projection Alternative To The Softmax Function Highlight: In this work, we propose a method that adapts this parameter to individual training examples. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Weiwei Kong; Walid Krichene; Nicolas Mayoraz; Steffen Rendle; Li Zhang; | |
| 55 | Online Agnostic Boosting Via Regret Minimization Highlight: In this work we provide the first agnostic online boosting algorithm; that is, given a weak learner with only marginally-better-than-trivial regret guarantees, our algorithm boosts it to a strong learner with sublinear regret. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Nataly Brukhim; Xinyi Chen; Elad Hazan; Shay Moran; | |
| 56 | Causal Intervention For Weakly-Supervised Semantic Segmentation Highlight: We present a causal inference framework to improve Weakly-Supervised Semantic Segmentation (WSSS). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dong Zhang; Hanwang Zhang; Jinhui Tang; Xian-Sheng Hua; Qianru Sun; | |
| 57 | Belief Propagation Neural Networks Highlight: To bridge this gap, we introduce belief propagation neural networks (BPNNs), a class of parameterized operators that operate on factor graphs and generalize Belief Propagation (BP). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jonathan Kuck; Shuvam Chakraborty; Hao Tang; Rachel Luo; Jiaming Song; Ashish Sabharwal; Stefano Ermon; | |
| 58 | Over-parameterized Adversarial Training: An Analysis Overcoming The Curse Of Dimensionality Highlight: Our work proves convergence to low robust training loss for \emph{polynomial} width instead of exponential, under natural assumptions and with ReLU activations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yi Zhang; Orestis Plevrakis; Simon S. Du; Xingguo Li; Zhao Song; Sanjeev Arora; | |
| 59 | Post-training Iterative Hierarchical Data Augmentation For Deep Networks Highlight: In this paper, we propose a new iterative hierarchical data augmentation (IHDA) method to fine-tune trained deep neural networks to improve their generalization performance. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Adil Khan; Khadija Fraz; | |
| 60 | Debugging Tests For Model Explanations Highlight: We investigate whether post-hoc model explanations are effective for diagnosing model errors–model debugging. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Julius Adebayo; Michael Muelly; Ilaria Liccardi; Been Kim; | |
| 61 | Robust Compressed Sensing Using Generative Models Highlight: In this paper we propose an algorithm inspired by the Median-of-Means (MOM). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ajil Jalal; Liu Liu; Alexandros G. Dimakis; Constantine Caramanis; | |
| 62 | Fairness Without Demographics Through Adversarially Reweighted Learning Highlight: In this work we address this problem by proposing Adversarially Reweighted Learning (ARL). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Preethi Lahoti; Alex Beutel; Jilin Chen; Kang Lee; Flavien Prost; Nithum Thain; Xuezhi Wang; Ed Chi; | |
| 63 | Stochastic Latent Actor-Critic: Deep Reinforcement Learning With A Latent Variable Model Highlight: In this work, we tackle these two problems separately, by explicitly learning latent representations that can accelerate reinforcement learning from images. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Alex Lee; Anusha Nagabandi; Pieter Abbeel; Sergey Levine; | |
| 64 | Ridge Rider: Finding Diverse Solutions By Following Eigenvectors Of The Hessian Highlight: In this paper, we present a different approach. Rather than following the gradient, which corresponds to a locally greedy direction, we instead follow the eigenvectors of the Hessian. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jack Parker-Holder; Luke Metz; Cinjon Resnick; Hengyuan Hu; Adam Lerer; Alistair Letcher; Alexander Peysakhovich; Aldo Pacchiano; Jakob Foerster; | |
| 65 | The Route To Chaos In Routing Games: When Is Price Of Anarchy Too Optimistic? Highlight: We study MWU using the actual game costs without applying cost normalization to $[0,1]$. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Thiparat Chotibut; Fryderyk Falniowski; Michal Misiurewicz; Georgios Piliouras; | |
| 66 | Online Algorithm For Unsupervised Sequential Selection With Contextual Information Highlight: In this paper, we study Contextual Unsupervised Sequential Selection (USS), a new variant of the stochastic contextual bandits problem where the loss of an arm cannot be inferred from the observed feedback. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Arun Verma; Manjesh Kumar Hanawal; Csaba Szepesvari; Venkatesh Saligrama; | |
| 67 | Adapting Neural Architectures Between Domains Highlight: This paper aims to improve the generalization of neural architectures via domain adaptation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yanxi Li; zhaohui yang; Yunhe Wang; Chang Xu; | |
| 68 | What Went Wrong And When?\\ Instance-wise Feature Importance For Time-series Black-box Models Highlight: We propose FIT, a framework that evaluates the importance of observations for a multivariate time-series black-box model by quantifying the shift in the predictive distribution over time. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sana Tonekaboni; Shalmali Joshi; Kieran Campbell; David K. Duvenaud; Anna Goldenberg; | |
| 69 | Towards Better Generalization Of Adaptive Gradient Methods Highlight: To close this gap, we propose \textit{\textbf{S}table \textbf{A}daptive \textbf{G}radient \textbf{D}escent} (\textsc{SAGD}) for nonconvex optimization which leverages differential privacy to boost the generalization performance of adaptive gradient methods. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yingxue Zhou; Belhal Karimi; Jinxing Yu; Zhiqiang Xu; Ping Li; | |
| 70 | Learning Guidance Rewards With Trajectory-space Smoothing Highlight: This paper is in the same vein — starting with a surrogate RL objective that involves smoothing in the trajectory-space, we arrive at a new algorithm for learning guidance rewards. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tanmay Gangwani; Yuan Zhou; Jian Peng; | |
| 71 | Variance Reduction Via Accelerated Dual Averaging For Finite-Sum Optimization Highlight: In this paper, we introduce a simplified and unified method for finite-sum convex optimization, named \emph{Variance Reduction via Accelerated Dual Averaging (VRADA)}. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chaobing Song; Yong Jiang; Yi Ma; | |
| 72 | Tree! I Am No Tree! I Am A Low Dimensional Hyperbolic Embedding Highlight: In this paper, we explore a new method for learning hyperbolic representations by taking a metric-first approach. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Rishi Sonthalia; Anna Gilbert; | |
| 73 | Deep Structural Causal Models For Tractable Counterfactual Inference Highlight: We formulate a general framework for building structural causal models (SCMs) with deep learning components. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Nick Pawlowski; Daniel Coelho de Castro; Ben Glocker; | |
| 74 | Convolutional Generation Of Textured 3D Meshes Highlight: A key contribution of our work is the encoding of the mesh and texture as 2D representations, which are semantically aligned and can be easily modeled by a 2D convolutional GAN. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dario Pavllo; Graham Spinks; Thomas Hofmann; Marie-Francine Moens; Aurelien Lucchi; | |
| 75 | A Statistical Framework For Low-bitwidth Training Of Deep Neural Networks Highlight: In this paper, we address this problem by presenting a statistical framework for analyzing FQT algorithms. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jianfei Chen; Yu Gai; Zhewei Yao; Michael W. Mahoney; Joseph E. Gonzalez; | |
| 76 | Better Set Representations For Relational Reasoning Highlight: To resolve this limitation, we propose a simple and general network module called Set Refiner Network (SRN). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Qian Huang; Horace He; Abhay Singh; Yan Zhang; Ser Nam Lim; Austin R. Benson; | |
| 77 | AutoSync: Learning To Synchronize For Data-Parallel Distributed Deep Learning Highlight: In this paper, we develop a model- and resource-dependent representation for synchronization, which unifies multiple synchronization aspects ranging from architecture, message partitioning, placement scheme, to communication topology. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Hao Zhang; Yuan Li; Zhijie Deng; Xiaodan Liang; Lawrence Carin; Eric Xing; | |
| 78 | A Combinatorial Perspective On Transfer Learning Highlight: In this work we study how the learning of modular solutions can allow for effective generalization to both unseen and potentially differently distributed data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jianan Wang; Eren Sezener; David Budden; Marcus Hutter; Joel Veness; | |
| 79 | Hardness Of Learning Neural Networks With Natural Weights Highlight: We prove negative results in this regard, and show that for depth-$2$ networks, and many “natural" weights distributions such as the normal and the uniform distribution, most networks are hard to learn. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Amit Daniely; Gal Vardi; | |
| 80 | Higher-Order Spectral Clustering Of Directed Graphs Highlight: Based on the Hermitian matrix representation of digraphs, we present a nearly-linear time algorithm for digraph clustering, and further show that our proposed algorithm can be implemented in sublinear time under reasonable assumptions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Valdimar Steinar Ericsson Laenen; He Sun; | |
| 81 | Primal-Dual Mesh Convolutional Neural Networks Highlight: We propose a method that combines the advantages of both types of approaches, while addressing their limitations: we extend a primal-dual framework drawn from the graph-neural-network literature to triangle meshes, and define convolutions on two types of graphs constructed from an input mesh. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Francesco Milano; Antonio Loquercio; Antoni Rosinol; Davide Scaramuzza; Luca Carlone; | |
| 82 | The Advantage Of Conditional Meta-Learning For Biased Regularization And Fine Tuning Highlight: We address this limitation by conditional meta-learning, inferring a conditioning function mapping task’s side information into a meta-parameter vector that is appropriate for that task at hand. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Giulia Denevi; Massimiliano Pontil; Carlo Ciliberto; | |
| 83 | Watch Out! Motion Is Blurring The Vision Of Your Deep Neural Networks Highlight: We propose a novel adversarial attack method that can generate visually natural motion-blurred adversarial examples, named motion-based adversarial blur attack (ABBA). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Qing Guo; Felix Juefei-Xu; Xiaofei Xie; Lei Ma; Jian Wang; Bing Yu; Wei Feng; Yang Liu; | code |
| 84 | Sinkhorn Barycenter Via Functional Gradient Descent Highlight: In this paper, we consider the problem of computing the barycenter of a set of probability distributions under the Sinkhorn divergence. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zebang Shen; Zhenfu Wang; Alejandro Ribeiro; Hamed Hassani; | |
| 85 | Coresets For Near-Convex Functions Highlight: We suggest a generic framework for computing sensitivities (and thus coresets) for wide family of loss functions which we call near-convex functions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Morad Tukan; Alaa Maalouf; Dan Feldman; | |
| 86 | Bayesian Deep Ensembles Via The Neural Tangent Kernel Highlight: We introduce a simple modification to standard deep ensembles training, through addition of a computationally-tractable, randomised and untrainable function to each ensemble member, that enables a posterior interpretation in the infinite width limit. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Bobby He; Balaji Lakshminarayanan; Yee Whye Teh; | |
| 87 | Improved Schemes For Episodic Memory-based Lifelong Learning Highlight: In this paper, we provide the first unified view of episodic memory based approaches from an optimization’s perspective. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yunhui Guo; Mingrui Liu; Tianbao Yang; Tajana Rosing; | |
| 88 | Adaptive Sampling For Stochastic Risk-Averse Learning Highlight: We propose an adaptive sampling algorithm for stochastically optimizing the {\em Conditional Value-at-Risk (CVaR)} of a loss distribution, which measures its performance on the $\alpha$ fraction of most difficult examples. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sebastian Curi; Kfir Y. Levy; Stefanie Jegelka; Andreas Krause; | |
| 89 | Deep Wiener Deconvolution: Wiener Meets Deep Learning For Image Deblurring Highlight: We present a simple and effective approach for non-blind image deblurring, combining classical techniques and deep learning. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jiangxin Dong; Stefan Roth; Bernt Schiele; | |
| 90 | Discovering Reinforcement Learning Algorithms Highlight: This paper introduces a new meta-learning approach that discovers an entire update rule which includes both what to predict’ (e.g. value functions) and how to learn from it’ (e.g. bootstrapping) by interacting with a set of environments. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Junhyuk Oh; Matteo Hessel; Wojciech M. Czarnecki; Zhongwen Xu; Hado P. van Hasselt; Satinder Singh; David Silver; | |
| 91 | Taming Discrete Integration Via The Boon Of Dimensionality Highlight: The key contribution of this work addresses this scalability challenge via an efficient reduction of discrete integration to model counting. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jeffrey Dudek; Dror Fried; Kuldeep S Meel; | |
| 92 | Blind Video Temporal Consistency Via Deep Video Prior Highlight: To address this issue, we present a novel and general approach for blind video temporal consistency. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chenyang Lei; Yazhou Xing; Qifeng Chen; | |
| 93 | Simplify And Robustify Negative Sampling For Implicit Collaborative Filtering Highlight: In this paper, we ?rst provide a novel understanding of negative instances by empirically observing that only a few instances are potentially important for model learning, and false negatives tend to have stable predictions over many training iterations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jingtao Ding; Yuhan Quan; Quanming Yao; Yong Li; Depeng Jin; | code |
| 94 | Model Selection For Production System Via Automated Online Experiments Highlight: We propose an automated online experimentation mechanism that can efficiently perform model selection from a large pool of models with a small number of online experiments. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zhenwen Dai; Praveen Chandar; Ghazal Fazelnia; Benjamin Carterette; Mounia Lalmas; | |
| 95 | On The Almost Sure Convergence Of Stochastic Gradient Descent In Non-Convex Problems Highlight: In this paper, we analyze the trajectories of stochastic gradient descent (SGD) with the aim of understanding their convergence properties in non-convex problems. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Panayotis Mertikopoulos; Nadav Hallak; Ali Kavis; Volkan Cevher; | |
| 96 | Automatic Perturbation Analysis For Scalable Certified Robustness And Beyond Highlight: In this paper, we develop an automatic framework to enable perturbation analysis on any neural network structures, by generalizing existing LiRPA algorithms such as CROWN to operate on general computational graphs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kaidi Xu; Zhouxing Shi; Huan Zhang; Yihan Wang; Kai-Wei Chang; Minlie Huang; Bhavya Kailkhura; Xue Lin; Cho-Jui Hsieh; | code |
| 97 | Adaptation Properties Allow Identification Of Optimized Neural Codes Highlight: Here we solve an inverse problem: characterizing the objective and constraint functions that efficient codes appear to be optimal for, on the basis of how they adapt to different stimulus distributions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Luke Rast; Jan Drugowitsch; | |
| 98 | Global Convergence And Variance Reduction For A Class Of Nonconvex-Nonconcave Minimax Problems Highlight: In this work, we show that for a subclass of nonconvex-nonconcave objectives satisfying a so-called two-sided Polyak-{\L}ojasiewicz inequality, the alternating gradient descent ascent (AGDA) algorithm converges globally at a linear rate and the stochastic AGDA achieves a sublinear rate. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Junchi Yang; Negar Kiyavash; Niao He; | |
| 99 | Model-Based Multi-Agent RL In Zero-Sum Markov Games With Near-Optimal Sample Complexity Highlight: In this paper, we aim to address the fundamental open question about the sample complexity of model-based MARL. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kaiqing Zhang; Sham Kakade; Tamer Basar; Lin Yang; | |
| 100 | Conservative Q-Learning For Offline Reinforcement Learning Highlight: In this paper, we propose conservative Q-learning (CQL), which aims to address these limitations by learning a conservative Q-function such that the expected value of a policy under this Q-function lower-bounds its true value. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Aviral Kumar; Aurick Zhou; George Tucker; Sergey Levine; | |
| 101 | Online Influence Maximization Under Linear Threshold Model Highlight: In this paper, we address OIM in the linear threshold (LT) model. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shuai Li; Fang Kong; Kejie Tang; Qizhi Li; Wei Chen; | |
| 102 | Ensembling Geophysical Models With Bayesian Neural Networks Highlight: We develop a novel data-driven ensembling strategy for combining geophysical models using Bayesian Neural Networks, which infers spatiotemporally varying model weights and bias while accounting for heteroscedastic uncertainties in the observations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ushnish Sengupta; Matt Amos; Scott Hosking; Carl Edward Rasmussen; Matthew Juniper; Paul Young; | |
| 103 | Delving Into The Cyclic Mechanism In Semi-supervised Video Object Segmentation Highlight: In this paper, we take attempt to incorporate the cyclic mechanism with the vision task of semi-supervised video object segmentation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yuxi Li; Ning Xu; Jinlong Peng; John See; Weiyao Lin; | |
| 104 | Asymmetric Shapley Values: Incorporating Causal Knowledge Into Model-agnostic Explainability Highlight: We introduce a less restrictive framework, Asymmetric Shapley values (ASVs), which are rigorously founded on a set of axioms, applicable to any AI system, and can flexibly incorporate any causal structure known to be respected by the data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Christopher Frye; Colin Rowat; Ilya Feige; | |
| 105 | Understanding Deep Architecture With Reasoning Layer Highlight: In this paper, we take an initial step toward an understanding of such hybrid deep architectures by showing that properties of the algorithm layers, such as convergence, stability and sensitivity, are intimately related to the approximation and generalization abilities of the end-to-end model. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Xinshi Chen; Yufei Zhang; Christoph Reisinger; Le Song; | |
| 106 | Planning In Markov Decision Processes With Gap-Dependent Sample Complexity Highlight: We propose MDP-GapE, a new trajectory-based Monte-Carlo Tree Search algorithm for planning in a Markov Decision Process in which transitions have a finite support. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Anders Jonsson; Emilie Kaufmann; Pierre Menard; Omar Darwiche Domingues; Edouard Leurent; Michal Valko; | |
| 107 | Provably Good Batch Off-Policy Reinforcement Learning Without Great Exploration Highlight: We show that using \emph{pessimistic value estimates} in the low-data regions in Bellman optimality and evaluation back-up can yield more adaptive and stronger guarantees when the concentrability assumption does not hold. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yao Liu; Adith Swaminathan; Alekh Agarwal; Emma Brunskill; | |
| 108 | Detection As Regression: Certified Object Detection With Median Smoothing Highlight: This work is motivated by recent progress on certified classification by randomized smoothing. We start by presenting a reduction from object detection to a regression problem. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ping-yeh Chiang; Michael Curry; Ahmed Abdelkader; Aounon Kumar; John Dickerson; Tom Goldstein; | |
| 109 | Contextual Reserve Price Optimization In Auctions Via Mixed Integer Programming Highlight: We study the problem of learning a linear model to set the reserve price in an auction, given contextual information, in order to maximize expected revenue from the seller side. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Joey Huchette; Haihao Lu; Hossein Esfandiari; Vahab Mirrokni; | |
| 110 | ExpandNets: Linear Over-parameterization To Train Compact Convolutional Networks Highlight: We introduce an approach to training a given compact network. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shuxuan Guo; Jose M. Alvarez; Mathieu Salzmann; | |
| 111 | FleXOR: Trainable Fractional Quantization Highlight: In this paper, we propose an encryption algorithm/architecture to compress quantized weights so as to achieve fractional numbers of bits per weight. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dongsoo Lee; Se Jung Kwon; Byeongwook Kim; Yongkweon Jeon; Baeseong Park; Jeongin Yun; | |
| 112 | The Implications Of Local Correlation On Learning Some Deep Functions Highlight: We introduce a property of distributions, denoted “local correlation”, which requires that small patches of the input image and of intermediate layers of the target function are correlated to the target label. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Eran Malach; Shai Shalev-Shwartz; | |
| 113 | Learning To Search Efficiently For Causally Near-optimal Treatments Highlight: We formalize this problem as learning a policy for finding a near-optimal treatment in a minimum number of trials using a causal inference framework. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Samuel H�kansson; Viktor Lindblom; Omer Gottesman; Fredrik D. Johansson; | |
| 114 | A Game Theoretic Analysis Of Additive Adversarial Attacks And Defenses Highlight: In this paper, we propose a game-theoretic framework for studying attacks and defenses which exist in equilibrium. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ambar Pal; Rene Vidal; | |
| 115 | Posterior Network: Uncertainty Estimation Without OOD Samples Via Density-Based Pseudo-Counts Highlight: In this work we propose the Posterior Network (PostNet), which uses Normalizing Flows to predict an individual closed-form posterior distribution over predicted probabilites for any input sample. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Bertrand Charpentier; Daniel Z�gner; Stephan G�nnemann; | |
| 116 | Recurrent Quantum Neural Networks Highlight: In this work we construct the first quantum recurrent neural network (QRNN) with demonstrable performance on non-trivial tasks such as sequence learning and integer digit classification. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Johannes Bausch; | |
| 117 | No-Regret Learning And Mixed Nash Equilibria: They Do Not Mix Highlight: In this paper, we study the dynamics of follow the regularized leader (FTRL), arguably the most well-studied class of no-regret dynamics, and we establish a sweeping negative result showing that the notion of mixed Nash equilibrium is antithetical to no-regret learning. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Emmanouil-Vasileios Vlatakis-Gkaragkounis; Lampros Flokas; Thanasis Lianeas; Panayotis Mertikopoulos; Georgios Piliouras; | |
| 118 | A Unifying View Of Optimism In Episodic Reinforcement Learning Highlight: In this paper we provide a general framework for designing, analyzing and implementing such algorithms in the episodic reinforcement learning problem. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Gergely Neu; Ciara Pike-Burke; | |
| 119 | Continuous Submodular Maximization: Beyond DR-Submodularity Highlight: In this paper, we propose the first continuous optimization algorithms that achieve a constant factor approximation guarantee for the problem of monotone continuous submodular maximization subject to a linear constraint. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Moran Feldman; Amin Karbasi; | |
| 120 | An Asymptotically Optimal Primal-Dual Incremental Algorithm For Contextual Linear Bandits Highlight: In this paper, we follow recent approaches of deriving asymptotically optimal algorithms from problem-dependent regret lower bounds and we introduce a novel algorithm improving over the state-of-the-art along multiple dimensions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Andrea Tirinzoni; Matteo Pirotta; Marcello Restelli; Alessandro Lazaric; | |
| 121 | Assessing SATNet's Ability To Solve The Symbol Grounding Problem Highlight: In this paper, we clarify SATNet’s capabilities by showing that in the absence of intermediate labels that identify individual Sudoku digit images with their logical representations, SATNet completely fails at visual Sudoku (0% test accuracy). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Oscar Chang; Lampros Flokas; Hod Lipson; Michael Spranger; | |
| 122 | A Bayesian Nonparametrics View Into Deep Representations Highlight: We investigate neural network representations from a probabilistic perspective. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Michal Jamroz; Marcin Kurdziel; Mateusz Opala; | |
| 123 | On The Similarity Between The Laplace And Neural Tangent Kernels Highlight: Here we show that NTK for fully connected networks with ReLU activation is closely related to the standard Laplace kernel. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Amnon Geifman; Abhay Yadav; Yoni Kasten; Meirav Galun; David Jacobs; Basri Ronen; | |
| 124 | A Causal View Of Compositional Zero-shot Recognition Highlight: Here we describe an approach for compositional generalization that builds on causal ideas. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yuval Atzmon; Felix Kreuk; Uri Shalit; Gal Chechik; | |
| 125 | HiPPO: Recurrent Memory With Optimal Polynomial Projections Highlight: We introduce a general framework (HiPPO) for the online compression of continuous signals and discrete time series by projection onto polynomial bases. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Albert Gu; Tri Dao; Stefano Ermon; Atri Rudra; Christopher R�; | |
| 126 | Auto Learning Attention Highlight: In this paper, we devise an Auto Learning Attention (AutoLA) method, which is the first attempt on automatic attention design. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Benteng Ma; Jing Zhang; Yong Xia; Dacheng Tao; | |
| 127 | CASTLE: Regularization Via Auxiliary Causal Graph Discovery Highlight: We introduce Causal Structure Learning (CASTLE) regularization and propose to regularize a neural network by jointly learning the causal relationships between variables. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Trent Kyono; Yao Zhang; Mihaela van der Schaar; | |
| 128 | Long-Tailed Classification By Keeping The Good And Removing The Bad Momentum Causal Effect Highlight: In this paper, we establish a causal inference framework, which not only unravels the whys of previous methods, but also derives a new principled solution. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kaihua Tang; Jianqiang Huang; Hanwang Zhang; | |
| 129 | Explainable Voting Highlight: We prove, however, that outcomes of the important Borda rule can be explained using O(m^2) steps, where m is the number of alternatives. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dominik Peters; Ariel D. Procaccia; Alexandros Psomas; Zixin Zhou; | |
| 130 | Deep Archimedean Copulas Highlight: In this paper, we introduce ACNet, a novel differentiable neural network architecture that enforces structural properties and enables one to learn an important class of copulas–Archimedean Copulas. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chun Kai Ling; Fei Fang; J. Zico Kolter; | |
| 131 | Re-Examining Linear Embeddings For High-Dimensional Bayesian Optimization Highlight: In this paper, we identify several crucial issues and misconceptions about the use of linear embeddings for BO. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ben Letham; Roberto Calandra; Akshara Rai; Eytan Bakshy; | |
| 132 | UnModNet: Learning To Unwrap A Modulo Image For High Dynamic Range Imaging Highlight: In this paper, we reformulate the modulo image unwrapping problem into a series of binary labeling problems and propose a modulo edge-aware model, named as UnModNet, to iteratively estimate the binary rollover masks of the modulo image for unwrapping. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chu Zhou; Hang Zhao; Jin Han; Chang Xu; Chao Xu; Tiejun Huang; Boxin Shi; | |
| 133 | Thunder: A Fast Coordinate Selection Solver For Sparse Learning Highlight: In this paper, we propose a novel active incremental approach to further improve the efficiency of the solvers. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shaogang Ren; Weijie Zhao; Ping Li; | |
| 134 | Neural Networks Fail To Learn Periodic Functions And How To Fix It Highlight: As a fix of this problem, we propose a new activation, namely, $x + \sin^2(x)$, which achieves the desired periodic inductive bias to learn a periodic function while maintaining a favorable optimization property of the $\relu$-based activations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Liu Ziyin; Tilman Hartwig; Masahito Ueda; | |
| 135 | Distribution Matching For Crowd Counting Highlight: In this paper, we show that imposing Gaussians to annotations hurts generalization performance. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Boyu Wang; Huidong Liu; Dimitris Samaras; Minh Hoai Nguyen; | code |
| 136 | Correspondence Learning Via Linearly-invariant Embedding Highlight: In this paper, we propose a fully differentiable pipeline for estimating accurate dense correspondences between 3D point clouds. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Riccardo Marin; Marie-Julie Rakotosaona; Simone Melzi; Maks Ovsjanikov; | |
| 137 | Learning To Dispatch For Job Shop Scheduling Via Deep Reinforcement Learning Highlight: In this paper, we propose to automatically learn PDRs via an end-to-end deep reinforcement learning agent. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Cong Zhang; Wen Song; Zhiguang Cao; Jie Zhang; Puay Siew Tan; Xu Chi; | |
| 138 | On Adaptive Attacks To Adversarial Example Defenses Highlight: While prior evaluation papers focused mainly on the end result—showing that a defense was ineffective—this paper focuses on laying out the methodology and the approach necessary to perform an adaptive attack. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Florian Tramer; Nicholas Carlini; Wieland Brendel; Aleksander Madry; | |
| 139 | Sinkhorn Natural Gradient For Generative Models Highlight: In this regard, we propose a novel Sinkhorn Natural Gradient (SiNG) algorithm which acts as a steepest descent method on the probability space endowed with the Sinkhorn divergence. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zebang Shen; Zhenfu Wang; Alejandro Ribeiro; Hamed Hassani; | |
| 140 | Online Sinkhorn: Optimal Transport Distances From Sample Streams Highlight: This paper introduces a new online estimator of entropy-regularized OT distances between two such arbitrary distributions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Arthur Mensch; Gabriel Peyr�; | |
| 141 | Ultrahyperbolic Representation Learning Highlight: In this paper, we propose a representation living on a pseudo-Riemannian manifold of constant nonzero curvature. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Marc Law; Jos Stam; | |
| 142 | Locally-Adaptive Nonparametric Online Learning Highlight: We fill this gap by introducing efficient online algorithms (based on a single versatile master algorithm) each adapting to one of the following regularities: (i) local Lipschitzness of the competitor function, (ii) local metric dimension of the instance sequence, (iii) local performance of the predictor across different regions of the instance space. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ilja Kuzborskij; Nicol� Cesa-Bianchi; | |
| 143 | Compositional Generalization Via Neural-Symbolic Stack Machines Highlight: To tackle this issue, we propose the Neural-Symbolic Stack Machine (NeSS). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Xinyun Chen; Chen Liang; Adams Wei Yu; Dawn Song; Denny Zhou; | |
| 144 | Graphon Neural Networks And The Transferability Of Graph Neural Networks Highlight: In this paper we introduce graphon NNs as limit objects of GNNs and prove a bound on the difference between the output of a GNN and its limit graphon-NN. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Luana Ruiz; Luiz Chamon; Alejandro Ribeiro; | |
| 145 | Unreasonable Effectiveness Of Greedy Algorithms In Multi-Armed Bandit With Many Arms Highlight: We study the structure of regret-minimizing policies in the {\em many-armed} Bayesian multi-armed bandit problem: in particular, with $k$ the number of arms and $T$ the time horizon, we consider the case where $k \geq \sqrt{T}$. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Mohsen Bayati; Nima Hamidi; Ramesh Johari; Khashayar Khosravi; | |
| 146 | Gamma-Models: Generative Temporal Difference Learning For Infinite-Horizon Prediction Highlight: We introduce the gamma-model, a predictive model of environment dynamics with an infinite, probabilistic horizon. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Michael Janner; Igor Mordatch; Sergey Levine; | |
| 147 | Deep Transformers With Latent Depth Highlight: We present a probabilistic framework to automatically learn which layer(s) to use by learning the posterior distributions of layer selection. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Xian Li; Asa Cooper Stickland; Yuqing Tang; Xiang Kong; | |
| 148 | Neural Mesh Flow: 3D Manifold Mesh Generation Via Diffeomorphic Flows Highlight: In this work, we propose NeuralMeshFlow (NMF) to generate two-manifold meshes for genus-0 shapes. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kunal Gupta; Manmohan Chandraker; | |
| 149 | Statistical Control For Spatio-temporal MEG/EEG Source Imaging With Desparsified Mutli-task Lasso Highlight: To deal with this, we adapt the desparsified Lasso estimator —an estimator tailored for high dimensional linear model that asymptotically follows a Gaussian distribution under sparsity and moderate feature correlation assumptions— to temporal data corrupted with autocorrelated noise. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jerome-Alexis Chevalier; Joseph Salmon; Alexandre Gramfort; Bertrand Thirion; | |
| 150 | A Scalable MIP-based Method For Learning Optimal Multivariate Decision Trees Highlight: In this paper, we propose a novel MIP formulation, based on 1-norm support vector machine model, to train a binary oblique ODT for classification problems. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Haoran Zhu; Pavankumar Murali; Dzung Phan; Lam Nguyen; Jayant Kalagnanam; | |
| 151 | Efficient Exact Verification Of Binarized Neural Networks Highlight: We present a new system, EEV, for efficient and exact verification of BNNs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kai Jia; Martin Rinard; | |
| 152 | Ultra-Low Precision 4-bit Training Of Deep Neural Networks Highlight: In this paper, we propose a number of novel techniques and numerical representation formats that enable, for the very first time, the precision of training systems to be aggressively scaled from 8-bits to 4-bits. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Xiao Sun; Naigang Wang; Chia-Yu Chen; Jiamin Ni; Ankur Agrawal; Xiaodong Cui; Swagath Venkataramani; Kaoutar El Maghraoui; Vijayalakshmi (Viji) Srinivasan; Kailash Gopalakrishnan; | |
| 153 | Bridging The Gap Between Sample-based And One-shot Neural Architecture Search With BONAS Highlight: In this work, we propose BONAS (Bayesian Optimized Neural Architecture Search), a sample-based NAS framework which is accelerated using weight-sharing to evaluate multiple related architectures simultaneously. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Han Shi; Renjie Pi; Hang Xu; Zhenguo Li; James Kwok; Tong Zhang; | |
| 154 | On Numerosity Of Deep Neural Networks Highlight: Recently, a provocative claim was published that number sense spontaneously emerges in a deep neural network trained merely for visual object recognition. This has, if true, far reaching significance to the fields of machine learning and cognitive science alike. In this paper, we prove the above claim to be unfortunately incorrect. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Xi Zhang; Xiaolin Wu; | |
| 155 | Outlier Robust Mean Estimation With Subgaussian Rates Via Stability Highlight: We study the problem of outlier robust high-dimensional mean estimation under a bounded covariance assumption, and more broadly under bounded low-degree moment assumptions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ilias Diakonikolas; Daniel M. Kane; Ankit Pensia; | |
| 156 | Self-Supervised Relationship Probing Highlight: In this work, we introduce a self-supervised method that implicitly learns the visual relationships without relying on any ground-truth visual relationship annotations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jiuxiang Gu; Jason Kuen; Shafiq Joty; Jianfei Cai; Vlad Morariu; Handong Zhao; Tong Sun; | |
| 157 | Information Theoretic Counterfactual Learning From Missing-Not-At-Random Feedback Highlight: To circumvent the use of RCTs, we build an information theoretic counterfactual variational information bottleneck (CVIB), as an alternative for debiasing learning without RCTs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zifeng Wang; Xi Chen; Rui Wen; Shao-Lun Huang; Ercan Kuruoglu; Yefeng Zheng; | |
| 158 | Prophet Attention: Predicting Attention With Future Attention Highlight: In this paper, we propose the Prophet Attention, similar to the form of self-supervision. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Fenglin Liu; Xuancheng Ren; Xian Wu; Shen Ge; Wei Fan; Yuexian Zou; Xu Sun; | |
| 159 | Language Models Are Few-Shot Learners Highlight: Specifically, we train GPT-3, an autoregressive language model with 175 billion parameters, 10x more than any previous non-sparse language model, and test its performance in the few-shot setting. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tom Brown; Benjamin Mann; Nick Ryder; Melanie Subbiah; Jared D Kaplan; Prafulla Dhariwal; Arvind Neelakantan; Pranav Shyam; Girish Sastry; Amanda Askell; Sandhini Agarwal; Ariel Herbert-Voss; Gretchen Krueger; Tom Henighan; Rewon Child; Aditya Ramesh; Daniel Ziegler; Jeffrey Wu; Clemens Winter; Chris Hesse; Mark Chen; Eric Sigler; Mateusz Litwin; Scott Gray; Benjamin Chess; Jack Clark; Christopher Berner; Sam McCandlish; Alec Radford; Ilya Sutskever; Dario Amodei; | |
| 160 | Margins Are Insufficient For Explaining Gradient Boosting Highlight: In this work, we first demonstrate that the k’th margin bound is inadequate in explaining the performance of state-of-the-art gradient boosters. We then explain the short comings of the k’th margin bound and prove a stronger and more refined margin-based generalization bound that indeed succeeds in explaining the performance of modern gradient boosters. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Allan Gr�nlund; Lior Kamma; Kasper Green Larsen; | |
| 161 | Fourier-transform-based Attribution Priors Improve The Interpretability And Stability Of Deep Learning Models For Genomics Highlight: To address these shortcomings, we propose a novel attribution prior, where the Fourier transform of input-level attribution scores are computed at training-time, and high-frequency components of the Fourier spectrum are penalized. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Alex Tseng; Avanti Shrikumar; Anshul Kundaje; | |
| 162 | MomentumRNN: Integrating Momentum Into Recurrent Neural Networks Highlight: We theoretically prove and numerically demonstrate that MomentumRNNs alleviate the vanishing gradient issue in training RNNs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tan Nguyen; Richard Baraniuk; Andrea Bertozzi; Stanley Osher; Bao Wang; | |
| 163 | Marginal Utility For Planning In Continuous Or Large Discrete Action Spaces Highlight: In this paper we explore explicitly learning a candidate action generator by optimizing a novel objective, marginal utility. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zaheen Ahmad; Levi Lelis; Michael Bowling; | |
| 164 | Projected Stein Variational Gradient Descent Highlight: In this work, we propose a {projected Stein variational gradient descent} (pSVGD) method to overcome this challenge by exploiting the fundamental property of intrinsic low dimensionality of the data informed subspace stemming from ill-posedness of such problems. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Peng Chen; Omar Ghattas; | |
| 165 | Minimax Lower Bounds For Transfer Learning With Linear And One-hidden Layer Neural Networks Highlight: In this paper we develop a statistical minimax framework to characterize the fundamental limits of transfer learning in the context of regression with linear and one-hidden layer neural network models. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Mohammadreza Mousavi Kalan; Zalan Fabian; Salman Avestimehr; Mahdi Soltanolkotabi; | |
| 166 | SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks Highlight: We introduce the SE(3)-Transformer, a variant of the self-attention module for 3D point-clouds, which is equivariant under continuous 3D roto-translations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Fabian Fuchs; Daniel Worrall; Volker Fischer; Max Welling; | |
| 167 | On The Equivalence Of Molecular Graph Convolution And Molecular Wave Function With Poor Basis Set Highlight: In this study, we demonstrate that the linear combination of atomic orbitals (LCAO), an approximation introduced by Pauling and Lennard-Jones in the 1920s, corresponds to graph convolutional networks (GCNs) for molecules. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Masashi Tsubaki; Teruyasu Mizoguchi; | |
| 168 | The Power Of Predictions In Online Control Highlight: We study the impact of predictions in online Linear Quadratic Regulator control with both stochastic and adversarial disturbances in the dynamics. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chenkai Yu; Guanya Shi; Soon-Jo Chung; Yisong Yue; Adam Wierman; | |
| 169 | Learning Affordance Landscapes For Interaction Exploration In 3D Environments Highlight: We introduce a reinforcement learning approach for exploration for interaction, whereby an embodied agent autonomously discovers the affordance landscape of a new unmapped 3D environment (such as an unfamiliar kitchen). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tushar Nagarajan; Kristen Grauman; | code |
| 170 | Cooperative Multi-player Bandit Optimization Highlight: We design a distributed learning algorithm that overcomes the informational bias players have towards maximizing the rewards of nearby players they got more information about. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ilai Bistritz; Nicholas Bambos; | |
| 171 | Tight First- And Second-Order Regret Bounds For Adversarial Linear Bandits Highlight: We propose novel algorithms with first- and second-order regret bounds for adversarial linear bandits. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shinji Ito; Shuichi Hirahara; Tasuku Soma; Yuichi Yoshida; | |
| 172 | Just Pick A Sign: Optimizing Deep Multitask Models With Gradient Sign Dropout Highlight: We present Gradient Sign Dropout (GradDrop), a probabilistic masking procedure which samples gradients at an activation layer based on their level of consistency. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zhao Chen; Jiquan Ngiam; Yanping Huang; Thang Luong; Henrik Kretzschmar; Yuning Chai; Dragomir Anguelov; | |
| 173 | A Loss Function For Generative Neural Networks Based On Watson�s Perceptual Model Highlight: We propose such a loss function based on Watson’s perceptual model, which computes a weighted distance in frequency space and accounts for luminance and contrast masking. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Steffen Czolbe; Oswin Krause; Ingemar Cox; Christian Igel; | |
| 174 | Dynamic Fusion Of Eye Movement Data And Verbal Narrations In Knowledge-rich Domains Highlight: We propose to jointly analyze experts’ eye movements and verbal narrations to discover important and interpretable knowledge patterns to better understand their decision-making processes. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ervine Zheng; Qi Yu; Rui Li; Pengcheng Shi; Anne Haake; | |
| 175 | Scalable Multi-Agent Reinforcement Learning For Networked Systems With Average Reward Highlight: In this paper, we identify a rich class of networked MARL problems where the model exhibits a local dependence structure that allows it to be solved in a scalable manner. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Guannan Qu; Yiheng Lin; Adam Wierman; Na Li; | |
| 176 | Optimizing Neural Networks Via Koopman Operator Theory Highlight: Koopman operator theory, a powerful framework for discovering the underlying dynamics of nonlinear dynamical systems, was recently shown to be intimately connected with neural network training. In this work, we take the first steps in making use of this connection. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Akshunna Dogra; William Redman; | |
| 177 | SVGD As A Kernelized Wasserstein Gradient Flow Of The Chi-squared Divergence Highlight: We introduce a new perspective on SVGD that instead views SVGD as the kernelized gradient flow of the chi-squared divergence. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sinho Chewi; Thibaut Le Gouic; Chen Lu; Tyler Maunu; Philippe Rigollet; | |
| 178 | Adversarial Robustness Of Supervised Sparse Coding Highlight: In this work, we strike a better balance by considering a model that involves learning a representation while at the same time giving a precise generalization bound and a robustness certificate. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jeremias Sulam; Ramchandran Muthukumar; Raman Arora; | |
| 179 | Differentiable Meta-Learning Of Bandit Policies Highlight: In this work, we learn such policies for an unknown distribution P using samples from P. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Craig Boutilier; Chih-wei Hsu; Branislav Kveton; Martin Mladenov; Csaba Szepesvari; Manzil Zaheer; | |
| 180 | Biologically Inspired Mechanisms For Adversarial Robustness Highlight: In this work, we investigate the role of two biologically plausible mechanisms in adversarial robustness. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Manish Vuyyuru Reddy; Andrzej Banburski; Nishka Pant; Tomaso Poggio; | |
| 181 | Statistical-Query Lower Bounds Via Functional Gradients Highlight: For the specific problem of ReLU regression (equivalently, agnostically learning a ReLU), we show that any statistical-query algorithm with tolerance $n^{-(1/\epsilon)^b}$ must use at least $2^{n^c} \epsilon$ queries for some constants $b, c > 0$, where $n$ is the dimension and $\epsilon$ is the accuracy parameter. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Surbhi Goel; Aravind Gollakota; Adam Klivans; | |
| 182 | Near-Optimal Reinforcement Learning With Self-Play Highlight: This paper closes this gap for the first time: we propose an optimistic variant of the Nash Q-learning algorithm with sample complexity \tlO(SAB), and a new Nash V-learning algorithm with sample complexity \tlO(S(A+B)). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yu Bai; Chi Jin; Tiancheng Yu; | |
| 183 | Network Diffusions Via Neural Mean-Field Dynamics Highlight: We propose a novel learning framework based on neural mean-field dynamics for inference and estimation problems of diffusion on networks. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shushan He; Hongyuan Zha; Xiaojing Ye; | |
| 184 | Self-Distillation As Instance-Specific Label Smoothing Highlight: With this in mind, we offer a new interpretation for teacher-student training as amortized MAP estimation, such that teacher predictions enable instance-specific regularization. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zhilu Zhang; Mert Sabuncu; | |
| 185 | Towards Problem-dependent Optimal Learning Rates Highlight: In this paper we propose a new framework based on a "uniform localized convergence" principle. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yunbei Xu; Assaf Zeevi; | |
| 186 | Cross-lingual Retrieval For Iterative Self-Supervised Training Highlight: In this work, we found that the cross-lingual alignment can be further improved by training seq2seq models on sentence pairs mined using their own encoder outputs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chau Tran; Yuqing Tang; Xian Li; Jiatao Gu; | |
| 187 | Rethinking Pooling In Graph Neural Networks Highlight: In this paper, we build upon representative GNNs and introduce variants that challenge the need for locality-preserving representations, either using randomization or clustering on the complement graph. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Diego Mesquita; Amauri Souza; Samuel Kaski; | |
| 188 | Pointer Graph Networks Highlight: Here we introduce Pointer Graph Networks (PGNs) which augment sets or graphs with additional inferred edges for improved model generalisation ability. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Petar Velickovic; Lars Buesing; Matthew Overlan; Razvan Pascanu; Oriol Vinyals; Charles Blundell; | |
| 189 | Gradient Regularized V-Learning For Dynamic Treatment Regimes Highlight: In this paper, we introduce Gradient Regularized V-learning (GRV), a novel method for estimating the value function of a DTR. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yao Zhang; Mihaela van der Schaar; | |
| 190 | Faster Wasserstein Distance Estimation With The Sinkhorn Divergence Highlight: In this work, we propose instead to estimate it with the Sinkhorn divergence, which is also built on entropic regularization but includes debiasing terms. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
L�na�c Chizat; Pierre Roussillon; Flavien L�ger; Fran�ois-Xavier Vialard; Gabriel Peyr�; | |
| 191 | Forethought And Hindsight In Credit Assignment Highlight: We address the problem of credit assignment in reinforcement learning and explore fundamental questions regarding the way in which an agent can best use additional computation to propagate new information, by planning with internal models of the world to improve its predictions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Veronica Chelu; Doina Precup; Hado P. van Hasselt; | |
| 192 | Robust Recursive Partitioning For Heterogeneous Treatment Effects With Uncertainty Quantification Highlight: This paper develops a new method for subgroup analysis, R2P, that addresses all these weaknesses. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Hyun-Suk Lee; Yao Zhang; William Zame; Cong Shen; Jang-Won Lee; Mihaela van der Schaar; | |
| 193 | Rescuing Neural Spike Train Models From Bad MLE Highlight: To alleviate this, we propose to directly minimize the divergence between neural recorded and model generated spike trains using spike train kernels. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Diego Arribas; Yuan Zhao; Il Memming Park; | |
| 194 | Lower Bounds And Optimal Algorithms For Personalized Federated Learning Highlight: In this work, we consider the optimization formulation of personalized federated learning recently introduced by Hanzely & Richtarik (2020) which was shown to give an alternative explanation to the workings of local SGD methods. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Filip Hanzely; Slavom�r Hanzely; Samuel Horv�th; Peter Richtarik; | |
| 195 | Black-Box Certification With Randomized Smoothing: A Functional Optimization Based Framework Highlight: We propose a general framework of adversarial certification with non-Gaussian noise and for more general types of attacks, from a unified \functional optimization perspective. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dinghuai Zhang; Mao Ye; Chengyue Gong; Zhanxing Zhu; Qiang Liu; | |
| 196 | Deep Imitation Learning For Bimanual Robotic Manipulation Highlight: We present a deep imitation learning framework for robotic bimanual manipulation in a continuous state-action space. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Fan Xie; Alexander Chowdhury; M. Clara De Paolis Kaluza; Linfeng Zhao; Lawson Wong; Rose Yu; | code |
| 197 | Stationary Activations For Uncertainty Calibration In Deep Learning Highlight: We introduce a new family of non-linear neural network activation functions that mimic the properties induced by the widely-used Mat\’ern family of kernels in Gaussian process (GP) models. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lassi Meronen; Christabella Irwanto; Arno Solin; | |
| 198 | Ensemble Distillation For Robust Model Fusion In Federated Learning Highlight: In this work we investigate more powerful and more flexible aggregation schemes for FL. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tao Lin; Lingjing Kong; Sebastian U. Stich; Martin Jaggi; | |
| 199 | Falcon: Fast Spectral Inference On Encrypted Data Highlight: In this paper, we propose a fast, frequency-domain deep neural network called Falcon, for fast inferences on encrypted data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Qian Lou; Wen-jie Lu; Cheng Hong; Lei Jiang; | |
| 200 | On Power Laws In Deep Ensembles Highlight: In this work, we focus on a classification problem and investigate the behavior of both non-calibrated and calibrated negative log-likelihood (CNLL) of a deep ensemble as a function of the ensemble size and the member network size. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ekaterina Lobacheva; Nadezhda Chirkova; Maxim Kodryan; Dmitry P. Vetrov; | |
| 201 | Practical Quasi-Newton Methods For Training Deep Neural Networks Highlight: We consider the development of practical stochastic quasi-Newton, and in particular Kronecker-factored block diagonal BFGS and L-BFGS methods, for training deep neural networks (DNNs). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Donald Goldfarb; Yi Ren; Achraf Bahamou; | |
| 202 | Approximation Based Variance Reduction For Reparameterization Gradients Highlight: In this work we present a control variate that is applicable for any reparameterizable distribution with known mean and covariance, e.g. Gaussians with any covariance structure. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tomas Geffner; Justin Domke; | |
| 203 | Inference Stage Optimization For Cross-scenario 3D Human Pose Estimation Highlight: In this work, we propose a novel framework, Inference Stage Optimization (ISO), for improving the generalizability of 3D pose models when source and target data come from different pose distributions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jianfeng Zhang; Xuecheng Nie; Jiashi Feng; | |
| 204 | Consistent Feature Selection For Analytic Deep Neural Networks Highlight: In this work, we investigate the problem of feature selection for analytic deep networks. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Vu C. Dinh; Lam S. Ho; | |
| 205 | Glance And Focus: A Dynamic Approach To Reducing Spatial Redundancy In Image Classification Highlight: Inspired by the fact that not all regions in an image are task-relevant, we propose a novel framework that performs efficient image classification by processing a sequence of relatively small inputs, which are strategically selected from the original image with reinforcement learning. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yulin Wang; Kangchen Lv; Rui Huang; Shiji Song; Le Yang; Gao Huang; | code |
| 206 | Information Maximization For Few-Shot Learning Highlight: We introduce Transductive Infomation Maximization (TIM) for few-shot learning. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Malik Boudiaf; Imtiaz Ziko; J�r�me Rony; Jose Dolz; Pablo Piantanida; Ismail Ben Ayed; | |
| 207 | Inverse Reinforcement Learning From A Gradient-based Learner Highlight: In this paper, we propose a new algorithm for this setting, in which the goal is to recover the reward function being optimized by an agent, given a sequence of policies produced during learning. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Giorgia Ramponi; Gianluca Drappo; Marcello Restelli; | |
| 208 | Bayesian Multi-type Mean Field Multi-agent Imitation Learning Highlight: In this paper, we proposed Bayesian multi-type mean field multi-agent imitation learning (BM3IL). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Fan Yang; Alina Vereshchaka; Changyou Chen; Wen Dong; | |
| 209 | Bayesian Robust Optimization For Imitation Learning Highlight: To provide a bridge between these two extremes, we propose Bayesian Robust Optimization for Imitation Learning (BROIL). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Daniel Brown; Scott Niekum; Marek Petrik; | |
| 210 | Multiview Neural Surface Reconstruction By Disentangling Geometry And Appearance Highlight: In this work we address the challenging problem of multiview 3D surface reconstruction. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lior Yariv; Yoni Kasten; Dror Moran; Meirav Galun; Matan Atzmon; Basri Ronen; Yaron Lipman; | |
| 211 | Riemannian Continuous Normalizing Flows Highlight: To overcome this problem, we introduce Riemannian continuous normalizing flows, a model which admits the parametrization of flexible probability measures on smooth manifolds by defining flows as the solution to ordinary differential equations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Emile Mathieu; Maximilian Nickel; | |
| 212 | Attention-Gated Brain Propagation: How The Brain Can Implement Reward-based Error Backpropagation Highlight: We demonstrate a biologically plausible reinforcement learning scheme for deep networks with an arbitrary number of layers. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Isabella Pozzi; Sander Bohte; Pieter Roelfsema; | |
| 213 | Asymptotic Guarantees For Generative Modeling Based On The Smooth Wasserstein Distance Highlight: In this work, we conduct a thorough statistical study of the minimum smooth Wasserstein estimators (MSWEs), first proving the estimator’s measurability and asymptotic consistency. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ziv Goldfeld; Kristjan Greenewald; Kengo Kato; | |
| 214 | Online Robust Regression Via SGD On The L1 Loss Highlight: In contrast, we show in this work that stochastic gradient descent on the l1 loss converges to the true parameter vector at a $\tilde{O}( 1 / (1 – \eta)^2 n )$ rate which is independent of the values of the contaminated measurements. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Scott Pesme; Nicolas Flammarion; | |
| 215 | PRANK: Motion Prediction Based On RANKing Highlight: In this paper, we introduce the PRANK method, which satisfies these requirements. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yuriy Biktairov; Maxim Stebelev; Irina Rudenko; Oleh Shliazhko; Boris Yangel; | |
| 216 | Fighting Copycat Agents In Behavioral Cloning From Observation Histories Highlight: To combat this "copycat problem", we propose an adversarial approach to learn a feature representation that removes excess information about the previous expert action nuisance correlate, while retaining the information necessary to predict the next action. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chuan Wen; Jierui Lin; Trevor Darrell; Dinesh Jayaraman; Yang Gao; | |
| 217 | Tight Nonparametric Convergence Rates For Stochastic Gradient Descent Under The Noiseless Linear Model Highlight: We analyze the convergence of single-pass, fixed step-size stochastic gradient descent on the least-square risk under this model. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Rapha�l Berthier; Francis Bach; Pierre Gaillard; | |
| 218 | Structured Prediction For Conditional Meta-Learning Highlight: In this work, we propose a new perspective on conditional meta-learning via structured prediction. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ruohan Wang; Yiannis Demiris; Carlo Ciliberto; | |
| 219 | Optimal Lottery Tickets Via Subset Sum: Logarithmic Over-Parameterization Is Sufficient Highlight: In this work, we close the gap and offer an exponential improvement to the over-parameterization requirement for the existence of lottery tickets. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ankit Pensia; Shashank Rajput; Alliot Nagle; Harit Vishwakarma; Dimitris Papailiopoulos; | |
| 220 | The Hateful Memes Challenge: Detecting Hate Speech In Multimodal Memes Highlight: This work proposes a new challenge set for multimodal classification, focusing on detecting hate speech in multimodal memes. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Douwe Kiela; Hamed Firooz; Aravind Mohan; Vedanuj Goswami; Amanpreet Singh; Pratik Ringshia; Davide Testuggine; | |
| 221 | Stochasticity Of Deterministic Gradient Descent: Large Learning Rate For Multiscale Objective Function Highlight: This article suggests that deterministic Gradient Descent, which does not use any stochastic gradient approximation, can still exhibit stochastic behaviors. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lingkai Kong; Molei Tao; | |
| 222 | Identifying Learning Rules From Neural Network Observables Highlight: It is an open question as to what specific experimental measurements would need to be made to determine whether any given learning rule is operative in a real biological system. In this work, we take a "virtual experimental" approach to this problem. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Aran Nayebi; Sanjana Srivastava; Surya Ganguli; Daniel L. Yamins; | |
| 223 | Optimal Approximation – Smoothness Tradeoffs For Soft-Max Functions Highlight: Our goal is to identify the optimal approximation-smoothness tradeoffs for different measures of approximation and smoothness. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Alessandro Epasto; Mohammad Mahdian; Vahab Mirrokni; Emmanouil Zampetakis; | |
| 224 | Weakly-Supervised Reinforcement Learning For Controllable Behavior Highlight: In this work, we introduce a framework for using weak supervision to automatically disentangle this semantically meaningful subspace of tasks from the enormous space of nonsensical "chaff" tasks. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lisa Lee; Ben Eysenbach; Russ R. Salakhutdinov; Shixiang (Shane) Gu; Chelsea Finn; | |
| 225 | Improving Policy-Constrained Kidney Exchange Via Pre-Screening Highlight: We propose both a greedy heuristic and a Monte Carlo tree search, which outperforms previous approaches, using experiments on both synthetic data and real kidney exchange data from the United Network for Organ Sharing. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Duncan McElfresh; Michael Curry; Tuomas Sandholm; John Dickerson; | |
| 226 | Learning Abstract Structure For Drawing By Efficient Motor Program Induction Highlight: We show that people spontaneously learn abstract drawing procedures that support generalization, and propose a model of how learners can discover these reusable drawing procedures. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lucas Tian; Kevin Ellis; Marta Kryven; Josh Tenenbaum; | |
| 227 | Why Do Deep Residual Networks Generalize Better Than Deep Feedforward Networks? — A Neural Tangent Kernel Perspective Highlight: This paper studies this fundamental problem in deep learning from a so-called neural tangent kernel” perspective. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kaixuan Huang; Yuqing Wang; Molei Tao; Tuo Zhao; | |
| 228 | Dual Instrumental Variable Regression Highlight: We present a novel algorithm for non-linear instrumental variable (IV) regression, DualIV, which simplifies traditional two-stage methods via a dual formulation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Krikamol Muandet; Arash Mehrjou; Si Kai Lee; Anant Raj; | |
| 229 | Stochastic Gradient Descent In Correlated Settings: A Study On Gaussian Processes Highlight: In this paper, we focus on the Gaussian process (GP) and take a step forward towards breaking the barrier by proving minibatch SGD converges to a critical point of the full loss function, and recovers model hyperparameters with rate $O(\frac{1}{K})$ up to a statistical error term depending on the minibatch size. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Hao Chen; Lili Zheng; Raed AL Kontar; Garvesh Raskutti; | |
| 230 | Interventional Few-Shot Learning Highlight: Thanks to it, we propose a novel FSL paradigm: Interventional Few-Shot Learning (IFSL). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zhongqi Yue; Hanwang Zhang; Qianru Sun; Xian-Sheng Hua; | code |
| 231 | Minimax Value Interval For Off-Policy Evaluation And Policy Optimization Highlight: We study minimax methods for off-policy evaluation (OPE) using value functions and marginalized importance weights. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Nan Jiang; Jiawei Huang; | |
| 232 | Biased Stochastic First-Order Methods For Conditional Stochastic Optimization And Applications In Meta Learning Highlight: For this special setting, we propose an accelerated algorithm called biased SpiderBoost (BSpiderBoost) that matches the lower bound complexity. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yifan Hu; Siqi Zhang; Xin Chen; Niao He; | |
| 233 | ShiftAddNet: A Hardware-Inspired Deep Network Highlight: This paper presented ShiftAddNet, whose main inspiration is drawn from a common practice in energy-efficient hardware implementation, that is, multiplication can be instead performed with additions and logical bit-shifts. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Haoran You; Xiaohan Chen; Yongan Zhang; Chaojian Li; Sicheng Li; Zihao Liu; Zhangyang Wang; Yingyan Lin; | code |
| 234 | Network-to-Network Translation With Conditional Invertible Neural Networks Highlight: Therefore, we seek a model that can relate between different existing representations and propose to solve this task with a conditionally invertible network. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Robin Rombach; Patrick Esser; Bjorn Ommer; | |
| 235 | Intra-Processing Methods For Debiasing Neural Networks Highlight: In this work, we initiate the study of a new paradigm in debiasing research, intra-processing, which sits between in-processing and post-processing methods. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yash Savani; Colin White; Naveen Sundar Govindarajulu; | code |
| 236 | Finding Second-Order Stationary Points Efficiently In Smooth Nonconvex Linearly Constrained Optimization Problems Highlight: This paper proposes two efficient algorithms for computing approximate second-order stationary points (SOSPs) of problems with generic smooth non-convex objective functions and generic linear constraints. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Songtao Lu; Meisam Razaviyayn; Bo Yang; Kejun Huang; Mingyi Hong; | |
| 237 | Model-based Policy Optimization With Unsupervised Model Adaptation Highlight: In this paper, we investigate how to bridge the gap between real and simulated data due to inaccurate model estimation for better policy optimization. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jian Shen; Han Zhao; Weinan Zhang; Yong Yu; | |
| 238 | Implicit Regularization And Convergence For Weight Normalization Highlight: Here, we study the weight normalization (WN) method \cite{salimans2016weight} and a variant called reparametrized projected gradient descent (rPGD) for overparametrized least squares regression and some more general loss functions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Xiaoxia Wu; Edgar Dobriban; Tongzheng Ren; Shanshan Wu; Zhiyuan Li; Suriya Gunasekar; Rachel Ward; Qiang Liu; | |
| 239 | Geometric All-way Boolean Tensor Decomposition Highlight: In this work, we presented a computationally efficient BTD algorithm, namely Geometric Expansion for all-order Tensor Factorization (GETF), that sequentially identifies the rank-1 basis components for a tensor from a geometric perspective. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Changlin Wan; Wennan Chang; Tong Zhao; Sha Cao; Chi Zhang; | |
| 240 | Modular Meta-Learning With Shrinkage Highlight: Here, we propose a meta-learning approach that obviates the need for this often sub-optimal hand-selection. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yutian Chen; Abram L. Friesen; Feryal Behbahani; Arnaud Doucet; David Budden; Matthew Hoffman; Nando de Freitas; | |
| 241 | A/B Testing In Dense Large-Scale Networks: Design And Inference Highlight: In this paper, we present a novel strategy for accurately estimating the causal effects of a class of treatments in a dense large-scale network. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Preetam Nandy; Kinjal Basu; Shaunak Chatterjee; Ye Tu; | |
| 242 | What Neural Networks Memorize And Why: Discovering The Long Tail Via Influence Estimation Highlight: In this work we design experiments to test the key ideas in this theory. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Vitaly Feldman; Chiyuan Zhang; | |
| 243 | Partially View-aligned Clustering Highlight: In this paper, we study one challenging issue in multi-view data clustering. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zhenyu Huang; Peng Hu; Joey Tianyi Zhou; Jiancheng Lv; Xi Peng; | |
| 244 | Partial Optimal Tranport With Applications On Positive-Unlabeled Learning Highlight: In this paper, we address the partial Wasserstein and Gromov-Wasserstein problems and propose exact algorithms to solve them. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Laetitia Chapel; Mokhtar Z. Alaya / Laboratoire LITIS; Universit� de Rouen Normandie; Gilles Gasso; | |
| 245 | Toward The Fundamental Limits Of Imitation Learning Highlight: In this paper, we focus on understanding the minimax statistical limits of IL in episodic Markov Decision Processes (MDPs). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Nived Rajaraman; Lin Yang; Jiantao Jiao; Kannan Ramchandran; | |
| 246 | Logarithmic Pruning Is All You Need Highlight: In this work, we remove the most limiting assumptions of this previous work while providing significantly tighter bounds: the overparameterized network only needs a logarithmic factor (in all variables but depth) number of neurons per weight of the target subnetwork. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Laurent Orseau; Marcus Hutter; Omar Rivasplata; | |
| 247 | Hold Me Tight! Influence Of Discriminative Features On Deep Network Boundaries Highlight: In this work, we borrow tools from the field of adversarial robustness, and propose a new perspective that relates dataset features to the distance of samples to the decision boundary. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Guillermo Ortiz-Jimenez; Apostolos Modas; Seyed-Mohsen Moosavi-Dezfooli; Pascal Frossard; | |
| 248 | Learning From Mixtures Of Private And Public Populations Highlight: Inspired by the above example, we consider a model in which the population $\cD$ is a mixture of two possibly distinct sub-populations: a private sub-population $\Dprv$ of private and sensitive data, and a public sub-population $\Dpub$ of data with no privacy concerns. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Raef Bassily; Shay Moran; Anupama Nandi; | |
| 249 | Adversarial Weight Perturbation Helps Robust Generalization Highlight: In this paper, we investigate the weight loss landscape from a new perspective, and identify a clear correlation between the flatness of weight loss landscape and robust generalization gap. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dongxian Wu; Shu-Tao Xia; Yisen Wang; | |
| 250 | Stateful Posted Pricing With Vanishing Regret Via Dynamic Deterministic Markov Decision Processes Highlight: In this paper, a rather general online problem called \emph{dynamic resource allocation with capacity constraints (DRACC)} is introduced and studied in the realm of posted price mechanisms. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yuval Emek; Ron Lavi; Rad Niazadeh; Yangguang Shi; | |
| 251 | Adversarial Self-Supervised Contrastive Learning Highlight: In this paper, we propose a novel adversarial attack for unlabeled data, which makes the model confuse the instance-level identities of the perturbed data samples. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Minseon Kim; Jihoon Tack; Sung Ju Hwang; | |
| 252 | Normalizing Kalman Filters For Multivariate Time Series Analysis Highlight: To this extent, we present a novel approach reconciling classical state space models with deep learning methods. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Emmanuel de B�zenac; Syama Sundar Rangapuram; Konstantinos Benidis; Michael Bohlke-Schneider; Richard Kurle; Lorenzo Stella; Hilaf Hasson; Patrick Gallinari; Tim Januschowski; | |
| 253 | Learning To Summarize With Human Feedback Highlight: In this work, we show that it is possible to significantly improve summary quality by training a model to optimize for human preferences. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Nisan Stiennon; Long Ouyang; Jeffrey Wu; Daniel Ziegler; Ryan Lowe; Chelsea Voss; Alec Radford; Dario Amodei; Paul F. Christiano; | |
| 254 | Fourier Spectrum Discrepancies In Deep Network Generated Images Highlight: In this paper, we present an analysis of the high-frequency Fourier modes of real and deep network generated images and show that deep network generated images share an observable, systematic shortcoming in replicating the attributes of these high-frequency modes. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tarik Dzanic; Karan Shah; Freddie Witherden; | |
| 255 | Lamina-specific Neuronal Properties Promote Robust, Stable Signal Propagation In Feedforward Networks Highlight: Specifically, we found that signal transformations, made by each layer of neurons on an input-driven spike signal, demodulate signal distortions introduced by preceding layers. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dongqi Han; Erik De Schutter; Sungho Hong; | |
| 256 | Learning Dynamic Belief Graphs To Generalize On Text-Based Games Highlight: In this work, we investigate how an agent can plan and generalize in text-based games using graph-structured representations learned end-to-end from raw text. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ashutosh Adhikari; Xingdi Yuan; Marc-Alexandre C�t�; Mikul� Zelinka; Marc-Antoine Rondeau; Romain Laroche; Pascal Poupart; Jian Tang; Adam Trischler; Will Hamilton; | |
| 257 | Triple Descent And The Two Kinds Of Overfitting: Where & Why Do They Appear? Highlight: In this paper, we show that despite their apparent similarity, these two scenarios are inherently different. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
St�phane d'Ascoli; Levent Sagun; Giulio Biroli; | |
| 258 | Multimodal Graph Networks For Compositional Generalization In Visual Question Answering Highlight: In this paper, we propose to tackle this challenge by employing neural factor graphs to induce a tighter coupling between concepts in different modalities (e.g. images and text). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Raeid Saqur; Karthik Narasimhan; | |
| 259 | Learning Graph Structure With A Finite-State Automaton Layer Highlight: In this work, we study the problem of learning to derive abstract relations from the intrinsic graph structure. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Daniel Johnson; Hugo Larochelle; Daniel Tarlow; | |
| 260 | A Universal Approximation Theorem Of Deep Neural Networks For Expressing Probability Distributions Highlight: This paper studies the universal approximation property of deep neural networks for representing probability distributions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yulong Lu; Jianfeng Lu; | |
| 261 | Unsupervised Object-centric Video Generation And Decomposition In 3D Highlight: We instead propose to model a video as the view seen while moving through a scene with multiple 3D objects and a 3D background. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Paul Henderson; Christoph H. Lampert; | |
| 262 | Domain Generalization For Medical Imaging Classification With Linear-Dependency Regularization Highlight: In this paper, we introduce a simple but effective approach to improve the generalization capability of deep neural networks in the field of medical imaging classification. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Haoliang Li; Yufei Wang; Renjie Wan; Shiqi Wang; Tie-Qiang Li; Alex Kot; | |
| 263 | Multi-label Classification: Do Hamming Loss And Subset Accuracy Really Conflict With Each Other? Highlight: This paper provides an attempt to fill up this gap by analyzing the learning guarantees of the corresponding learning algorithms on both SA and HL measures. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Guoqiang Wu; Jun Zhu; | |
| 264 | A Novel Automated Curriculum Strategy To Solve Hard Sokoban Planning Instances Highlight: We present a novel {\em automated} curriculum approach that dynamically selects from a pool of unlabeled training instances of varying task complexity guided by our {\em difficulty quantum momentum} strategy. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dieqiao Feng; Carla P. Gomes; Bart Selman; | |
| 265 | Causal Analysis Of Covid-19 Spread In Germany Highlight: In this work, we study the causal relations among German regions in terms of the spread of Covid-19 since the beginning of the pandemic, taking into account the restriction policies that were applied by the different federal states. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Atalanti Mastakouri; Bernhard Sch�lkopf; | |
| 266 | Locally Private Non-asymptotic Testing Of Discrete Distributions Is Faster Using Interactive Mechanisms Highlight: We find separation rates for testing multinomial or more general discrete distributions under the constraint of alpha-local differential privacy. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Thomas Berrett; Cristina Butucea; | |
| 267 | Adaptive Gradient Quantization For Data-Parallel SGD Highlight: We empirically observe that the statistics of gradients of deep models change during the training. Motivated by this observation, we introduce two adaptive quantization schemes, ALQ and AMQ. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Fartash Faghri; Iman Tabrizian; Ilia Markov; Dan Alistarh; Daniel M. Roy; Ali Ramezani-Kebrya; | |
| 268 | Finite Continuum-Armed Bandits Highlight: Focusing on a nonparametric setting, where the mean reward is an unknown function of a one-dimensional covariate, we propose an optimal strategy for this problem. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Solenne Gaucher; | |
| 269 | Removing Bias In Multi-modal Classifiers: Regularization By Maximizing Functional Entropies Highlight: To alleviate this shortcoming, we propose a novel regularization term based on the functional entropy. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Itai Gat; Idan Schwartz; Alexander Schwing; Tamir Hazan; | |
| 270 | Compact Task Representations As A Normative Model For Higher-order Brain Activity Highlight: More specifically, we focus on MDPs whose state is based on action-observation histories, and we show how to compress the state space such that unnecessary redundancy is eliminated, while task-relevant information is preserved. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Severin Berger; Christian K. Machens; | |
| 271 | Robust-Adaptive Control Of Linear Systems: Beyond Quadratic Costs Highlight: We consider the problem of robust and adaptive model predictive control (MPC) of a linear system, with unknown parameters that are learned along the way (adaptive), in a critical setting where failures must be prevented (robust). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Edouard Leurent; Odalric-Ambrym Maillard; Denis Efimov; | |
| 272 | Co-exposure Maximization In Online Social Networks Highlight: In this paper, we study the problem of allocating seed users to opposing campaigns: by drawing on the equal-time rule of political campaigning on traditional media, our goal is to allocate seed users to campaigners with the aim to maximize the expected number of users who are co-exposed to both campaigns. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sijing Tu; Cigdem Aslay; Aristides Gionis; | |
| 273 | UCLID-Net: Single View Reconstruction In Object Space Highlight: In this paper, we show that building a geometry preserving 3-dimensional latent space helps the network concurrently learn global shape regularities and local reasoning in the object coordinate space and, as a result, boosts performance. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Benoit Guillard; Edoardo Remelli; Pascal Fua; | |
| 274 | Reinforcement Learning For Control With Multiple Frequencies Highlight: In this paper, we formalize the problem of multiple control frequencies in RL and provide its efficient solution method. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jongmin Lee; ByungJun Lee; Kee-Eung Kim; | |
| 275 | Complex Dynamics In Simple Neural Networks: Understanding Gradient Flow In Phase Retrieval Highlight: Here we focus on gradient flow dynamics for phase retrieval from random measurements. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Stefano Sarao Mannelli; Giulio Biroli; Chiara Cammarota; Florent Krzakala; Pierfrancesco Urbani; Lenka Zdeborov�; | |
| 276 | Neural Message Passing For Multi-Relational Ordered And Recursive Hypergraphs Highlight: In this work, we first unify exisiting MPNNs on different structures into G-MPNN (Generalised MPNN) framework. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Naganand Yadati; | |
| 277 | A Unified View Of Label Shift Estimation Highlight: In this paper, we present a unified view of the two methods and the first theoretical characterization of MLLS. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Saurabh Garg; Yifan Wu; Sivaraman Balakrishnan; Zachary Lipton; | |
| 278 | Optimal Private Median Estimation Under Minimal Distributional Assumptions Highlight: We study the fundamental task of estimating the median of an underlying distribution from a finite number of samples, under pure differential privacy constraints. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Christos Tzamos; Emmanouil-Vasileios Vlatakis-Gkaragkounis; Ilias Zadik; | |
| 279 | Breaking The Communication-Privacy-Accuracy Trilemma Highlight: In this paper, we develop novel encoding and decoding mechanisms that simultaneously achieve optimal privacy and communication efficiency in various canonical settings. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Wei-Ning Chen; Peter Kairouz; Ayfer Ozgur; | |
| 280 | Audeo: Audio Generation For A Silent Performance Video Highlight: Our main aim in this work is to explore the plausibility of such a transformation and to identify cues and components able to carry the association of sounds with visual events. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kun Su; Xiulong Liu; Eli Shlizerman; | |
| 281 | Ode To An ODE Highlight: We present a new paradigm for Neural ODE algorithms, called ODEtoODE, where time-dependent parameters of the main flow evolve according to a matrix flow on the orthogonal group O(d). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Krzysztof M. Choromanski; Jared Quincy Davis; Valerii Likhosherstov; Xingyou Song; Jean-Jacques Slotine; Jacob Varley; Honglak Lee; Adrian Weller; Vikas Sindhwani; | |
| 282 | Self-Distillation Amplifies Regularization In Hilbert Space Highlight: This work provides the first theoretical analysis of self-distillation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Hossein Mobahi; Mehrdad Farajtabar; Peter Bartlett; | |
| 283 | Coupling-based Invertible Neural Networks Are Universal Diffeomorphism Approximators Highlight: Without a universality, there could be a well-behaved invertible transformation that the CF-INN can never approximate, hence it would render the model class unreliable. We answer this question by showing a convenient criterion: a CF-INN is universal if its layers contain affine coupling and invertible linear functions as special cases. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Takeshi Teshima; Isao Ishikawa; Koichi Tojo; Kenta Oono; Masahiro Ikeda; Masashi Sugiyama; | |
| 284 | Community Detection Using Fast Low-cardinality Semidefinite Programming? Highlight: In this paper, we propose a new class of low-cardinality algorithm that generalizes the local update to maximize a semidefinite relaxation derived from max-k-cut. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Po-Wei Wang; J. Zico Kolter; | |
| 285 | Modeling Noisy Annotations For Crowd Counting Highlight: In this paper, we first model the annotation noise using a random variable with Gaussian distribution, and derive the pdf of the crowd density value for each spatial location in the image. We then approximate the joint distribution of the density values (i.e., the distribution of density maps) with a full covariance multivariate Gaussian density, and derive a low-rank approximate for tractable implementation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jia Wan; Antoni Chan; | |
| 286 | An Operator View Of Policy Gradient Methods Highlight: We cast policy gradient methods as the repeated application of two operators: a policy improvement operator $\mathcal{I}$, which maps any policy $\pi$ to a better one $\mathcal{I}\pi$, and a projection operator $\mathcal{P}$, which finds the best approximation of $\mathcal{I}\pi$ in the set of realizable policies. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dibya Ghosh; Marlos C. Machado; Nicolas Le Roux; | |
| 287 | Demystifying Contrastive Self-Supervised Learning: Invariances, Augmentations And Dataset Biases Highlight: Somewhat mysteriously the recent gains in performance come from training instance classification models, treating each image and it’s augmented versions as samples of a single class. In this work, we first present quantitative experiments to demystify these gains. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Senthil Purushwalkam Shiva Prakash; Abhinav Gupta; | |
| 288 | Online MAP Inference Of Determinantal Point Processes Highlight: In this paper, we provide an efficient approximation algorithm for finding the most likelihood configuration (MAP) of size $k$ for Determinantal Point Processes (DPP) in the online setting where the data points arrive in an arbitrary order and the algorithm cannot discard the selected elements from its local memory. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Aditya Bhaskara; Amin Karbasi; Silvio Lattanzi; Morteza Zadimoghaddam; | |
| 289 | Video Object Segmentation With Adaptive Feature Bank And Uncertain-Region Refinement Highlight: This paper presents a new matching-based framework for semi-supervised video object segmentation (VOS). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yongqing Liang; Xin Li; Navid Jafari; Jim Chen; | |
| 290 | Inferring Learning Rules From Animal Decision-making Highlight: Whereas reinforcement learning often focuses on the design of algorithms that enable artificial agents to efficiently learn new tasks, here we develop a modeling framework to directly infer the empirical learning rules that animals use to acquire new behaviors. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zoe Ashwood; Nicholas A. Roy; Ji Hyun Bak; Jonathan W. Pillow; | |
| 291 | Input-Aware Dynamic Backdoor Attack Highlight: In this work, we propose a novel backdoor attack technique in which the triggers vary from input to input. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tuan Anh Nguyen; Anh Tran; | |
| 292 | How Hard Is To Distinguish Graphs With Graph Neural Networks? Highlight: This study derives hardness results for the classification variant of graph isomorphism in the message-passing model (MPNN). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Andreas Loukas; | |
| 293 | Minimax Regret Of Switching-Constrained Online Convex Optimization: No Phase Transition Highlight: In this paper, we show that $ T $-round switching-constrained OCO with fewer than $ K $ switches has a minimax regret of $ \Theta(\frac{T}{\sqrt{K}}) $. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lin Chen; Qian Yu; Hannah Lawrence; Amin Karbasi; | |
| 294 | Dual Manifold Adversarial Robustness: Defense Against Lp And Non-Lp Adversarial Attacks Highlight: To partially answer this question, we consider the scenario when the manifold information of the underlying data is available. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Wei-An Lin; Chun Pong Lau; Alexander Levine; Rama Chellappa; Soheil Feizi; | |
| 295 | Cross-Scale Internal Graph Neural Network For Image Super-Resolution Highlight: In this paper, we explore the cross-scale patch recurrence property of a natural image, i.e., similar patches tend to recur many times across different scales. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shangchen Zhou; Jiawei Zhang; Wangmeng Zuo; Chen Change Loy; | |
| 296 | Unsupervised Representation Learning By Invariance Propagation Highlight: In this paper, we propose Invariance Propagation to focus on learning representations invariant to category-level variations, which are provided by different instances from the same category. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Feng Wang; Huaping Liu; Di Guo; Sun Fuchun; | |
| 297 | Restoring Negative Information In Few-Shot Object Detection Highlight: In this paper, we restore the negative information in few-shot object detection by introducing a new negative- and positive-representative based metric learning framework and a new inference scheme with negative and positive representatives. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yukuan Yang; Fangyun Wei; Miaojing Shi; Guoqi Li; | code |
| 298 | Do Adversarially Robust ImageNet Models Transfer Better? Highlight: In this work, we identify another such aspect: we find that adversarially robust models, while less accurate, often perform better than their standard-trained counterparts when used for transfer learning. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Hadi Salman; Andrew Ilyas; Logan Engstrom; Ashish Kapoor; Aleksander Madry; | |
| 299 | Robust Correction Of Sampling Bias Using Cumulative Distribution Functions Highlight: We present a new method for handling covariate shift using the empirical cumulative distribution function estimates of the target distribution by a rigorous generalization of a recent idea proposed by Vapnik and Izmailov. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Bijan Mazaheri; Siddharth Jain; Jehoshua Bruck; | |
| 300 | Personalized Federated Learning With Theoretical Guarantees: A Model-Agnostic Meta-Learning Approach Highlight: In this paper, we study a personalized variant of the federated learning in which our goal is to find an initial shared model that current or new users can easily adapt to their local dataset by performing one or a few steps of gradient descent with respect to their own data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Alireza Fallah; Aryan Mokhtari; Asuman Ozdaglar; | |
| 301 | Pixel-Level Cycle Association: A New Perspective For Domain Adaptive Semantic Segmentation Highlight: In this paper, we propose to build the pixel-level cycle association between source and target pixel pairs and contrastively strengthen their connections to diminish the domain gap and make the features more discriminative. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Guoliang Kang; Yunchao Wei; Yi Yang; Yueting Zhuang; Alexander Hauptmann; | code |
| 302 | Classification With Valid And Adaptive Coverage Highlight: In this paper, we develop specialized versions of these techniques for categorical and unordered response labels that, in addition to providing marginal coverage, are also fully adaptive to complex data distributions, in the sense that they perform favorably in terms of approximate conditional coverage compared to alternative methods. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yaniv Romano; Matteo Sesia; Emmanuel Candes; | |
| 303 | Learning Global Transparent Models Consistent With Local Contrastive Explanations Highlight: In this work, we explore the question: Can we produce a transparent global model that is simultaneously accurate and consistent with the local (contrastive) explanations of the black-box model? Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tejaswini Pedapati; Avinash Balakrishnan; Karthikeyan Shanmugam; Amit Dhurandhar; | |
| 304 | Learning To Approximate A Bregman Divergence Highlight: In this paper, we focus on the problem of approximating an arbitrary Bregman divergence from supervision, and we provide a well-principled approach to analyzing such approximations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ali Siahkamari; XIDE XIA; Venkatesh Saligrama; David Casta��n; Brian Kulis; | |
| 305 | Diverse Image Captioning With Context-Object Split Latent Spaces Highlight: To this end, we introduce a novel factorization of the latent space, termed context-object split, to model diversity in contextual descriptions across images and texts within the dataset. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shweta Mahajan; Stefan Roth; | |
| 306 | Learning Disentangled Representations Of Videos With Missing Data Highlight: We present Disentangled Imputed Video autoEncoder (DIVE), a deep generative model that imputes and predicts future video frames in the presence of missing data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Armand Comas; Chi Zhang; Zlatan Feric; Octavia Camps; Rose Yu; | code |
| 307 | Natural Graph Networks Highlight: Here we show that instead of equivariance, the more general concept of naturality is sufficient for a graph network to be well-defined, opening up a larger class of graph networks. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Pim de Haan; Taco S. Cohen; Max Welling; | |
| 308 | Continual Learning With Node-Importance Based Adaptive Group Sparse Regularization Highlight: We propose a novel regularization-based continual learning method, dubbed as Adaptive Group Sparsity based Continual Learning (AGS-CL), using two group sparsity-based penalties. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sangwon Jung; Hongjoon Ahn; Sungmin Cha; Taesup Moon; | |
| 309 | Towards Crowdsourced Training Of Large Neural Networks Using Decentralized Mixture-of-Experts Highlight: In this work, we propose Learning@home: a novel neural network training paradigm designed to handle large amounts of poorly connected participants. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Maksim Riabinin; Anton Gusev; | |
| 310 | Bidirectional Convolutional Poisson Gamma Dynamical Systems Highlight: Incorporating the natural document-sentence-word structure into hierarchical Bayesian modeling, we propose convolutional Poisson gamma dynamical systems (PGDS) that introduce not only word-level probabilistic convolutions, but also sentence-level stochastic temporal transitions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
wenchao chen; Chaojie Wang; Bo Chen; Yicheng Liu; Hao Zhang; Mingyuan Zhou; | |
| 311 | Deep Reinforcement And InfoMax Learning Highlight: To test that hypothesis, we introduce an objective based on Deep InfoMax (DIM) which trains the agent to predict the future by maximizing the mutual information between its internal representation of successive timesteps. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Bogdan Mazoure; Remi Tachet des Combes; Thang Long DOAN; Philip Bachman; R Devon Hjelm; | |
| 312 | On Ranking Via Sorting By Estimated Expected Utility Highlight: We provide an answer to this question in the form of a structural characterization of ranking losses for which a suitable regression is consistent. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Clement Calauzenes; Nicolas Usunier; | |
| 313 | Distribution-free Binary Classification: Prediction Sets, Confidence Intervals And Calibration Highlight: We study three notions of uncertainty quantification—calibration, confidence intervals and prediction sets—for binary classification in the distribution-free setting, that is without making any distributional assumptions on the data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chirag Gupta; Aleksandr Podkopaev; Aaditya Ramdas; | |
| 314 | Closing The Dequantization Gap: PixelCNN As A Single-Layer Flow Highlight: In this paper, we introduce subset flows, a class of flows that can tractably transform finite volumes and thus allow exact computation of likelihoods for discrete data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Didrik Nielsen; Ole Winther; | |
| 315 | Sequence To Multi-Sequence Learning Via Conditional Chain Mapping For Mixture Signals Highlight: In this work, we focus on one-to-many sequence transduction problems, such as extracting multiple sequential sources from a mixture sequence. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jing Shi; Xuankai Chang; Pengcheng Guo; Shinji Watanabe; Yusuke Fujita; Jiaming Xu; Bo Xu; Lei Xie; | |
| 316 | Variance Reduction For Random Coordinate Descent-Langevin Monte Carlo Highlight: We show by a counterexamplethat blindly applying RCD does not achieve the goal in the most general setting. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
ZHIYAN DING; Qin Li; | |
| 317 | Language As A Cognitive Tool To Imagine Goals In Curiosity Driven Exploration Highlight: We introduce IMAGINE, an intrinsically motivated deep reinforcement learning architecture that models this ability. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
C�dric Colas; Tristan Karch; Nicolas Lair; Jean-Michel Dussoux; Cl�ment Moulin-Frier; Peter Dominey; Pierre-Yves Oudeyer; | |
| 318 | All Word Embeddings From One Embedding Highlight: In this study, to reduce the total number of parameters, the embeddings for all words are represented by transforming a shared embedding. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sho Takase; Sosuke Kobayashi; | |
| 319 | Primal Dual Interpretation Of The Proximal Stochastic Gradient Langevin Algorithm Highlight: We consider the task of sampling with respect to a log concave probability distribution. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Adil SALIM; Peter Richtarik; | |
| 320 | How To Characterize The Landscape Of Overparameterized Convolutional Neural Networks Highlight: Specifically, we consider the loss landscape of an overparameterized convolutional neural network (CNN) in the continuous limit, where the numbers of channels/hidden nodes in the hidden layers go to infinity. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yihong Gu; Weizhong Zhang; Cong Fang; Jason D. Lee; Tong Zhang; | |
| 321 | On The Tightness Of Semidefinite Relaxations For Certifying Robustness To Adversarial Examples Highlight: In this paper, we describe a geometric technique that determines whether this SDP certificate is exact, meaning whether it provides both a lower-bound on the size of the smallest adversarial perturbation, as well as a globally optimal perturbation that attains the lower-bound. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Richard Zhang; | |
| 322 | Submodular Meta-Learning Highlight: In this paper, we introduce a discrete variant of the Meta-learning framework. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Arman Adibi; Aryan Mokhtari; Hamed Hassani; | |
| 323 | Rethinking Pre-training And Self-training Highlight: Our study reveals the generality and flexibility of self-training with three additional insights: 1) stronger data augmentation and more labeled data further diminish the value of pre-training, 2) unlike pre-training, self-training is always helpful when using stronger data augmentation, in both low-data and high-data regimes, and 3) in the case that pre-training is helpful, self-training improves upon pre-training. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Barret Zoph; Golnaz Ghiasi; Tsung-Yi Lin; Yin Cui; Hanxiao Liu; Ekin Dogus Cubuk; Quoc Le; | |
| 324 | Unsupervised Sound Separation Using Mixture Invariant Training Highlight: In this paper, we propose a completely unsupervised method, mixture invariant training (MixIT), that requires only single-channel acoustic mixtures. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Scott Wisdom; Efthymios Tzinis; Hakan Erdogan; Ron Weiss; Kevin Wilson; John Hershey; | |
| 325 | Adaptive Discretization For Model-Based Reinforcement Learning Highlight: We introduce the technique of adaptive discretization to design an efficient model-based episodic reinforcement learning algorithm in large (potentially continuous) state-action spaces. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sean Sinclair; Tianyu Wang; Gauri Jain; Siddhartha Banerjee; Christina Yu; | |
| 326 | CodeCMR: Cross-Modal Retrieval For Function-Level Binary Source Code Matching Highlight: This paper proposes an end-to-end cross-modal retrieval network for binary source code matching, which achieves higher accuracy and requires less expert experience. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zeping Yu; Wenxin Zheng; Jiaqi Wang; Qiyi Tang; Sen Nie; Shi Wu; | |
| 327 | On Warm-Starting Neural Network Training Highlight: In this work, we take a closer look at this empirical phenomenon and try to understand when and how it occurs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jordan Ash; Ryan P. Adams; | |
| 328 | DAGs With No Fears: A Closer Look At Continuous Optimization For Learning Bayesian Networks Highlight: Informed by the KKT conditions, a local search post-processing algorithm is proposed and shown to substantially and universally improve the structural Hamming distance of all tested algorithms, typically by a factor of 2 or more. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dennis Wei; Tian Gao; yue yu; | |
| 329 | OOD-MAML: Meta-Learning For Few-Shot Out-of-Distribution Detection And Classification Highlight: We propose a few-shot learning method for detecting out-of-distribution (OOD) samples from classes that are unseen during training while classifying samples from seen classes using only a few labeled examples. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Taewon Jeong; Heeyoung Kim; | |
| 330 | An Imitation From Observation Approach To Transfer Learning With Dynamics Mismatch Highlight: In this paper, we show that one existing solution to this transfer problem– grounded action transformation –is closely related to the problem of imitation from observation (IfO): learning behaviors that mimic the observations of behavior demonstrations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Siddharth Desai; Ishan Durugkar; Haresh Karnan; Garrett Warnell; Josiah Hanna; Peter Stone; | |
| 331 | Learning About Objects By Learning To Interact With Them Highlight: Taking inspiration from infants learning from their environment through play and interaction, we present a computational framework to discover objects and learn their physical properties along this paradigm of Learning from Interaction. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Martin Lohmann; Jordi Salvador; Aniruddha Kembhavi; Roozbeh Mottaghi; | |
| 332 | Learning Discrete Distributions With Infinite Support Highlight: We present a novel approach to estimating discrete distributions with (potentially) infinite support in the total variation metric. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Doron Cohen; Aryeh Kontorovich; Geo?rey Wolfer; | |
| 333 | Dissecting Neural ODEs Highlight: In this work we “open the box”, further developing the continuous-depth formulation with the aim of clarifying the influence of several design choices on the underlying dynamics. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Stefano Massaroli; Michael Poli; Jinkyoo Park; Atsushi Yamashita; edit Hajime Asama; | |
| 334 | Teaching A GAN What Not To Learn Highlight: In this paper, we approach the supervised GAN problem from a different perspective, one that is motivated by the philosophy of the famous Persian poet Rumi who said, "The art of knowing is knowing what to ignore." Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Siddarth Asokan; Chandra Seelamantula; | |
| 335 | Counterfactual Data Augmentation Using Locally Factored Dynamics Highlight: We propose an approach to inferring these structures given an object-oriented state representation, as well as a novel algorithm for Counterfactual Data Augmentation (CoDA). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Silviu Pitis; Elliot Creager; Animesh Garg; | code |
| 336 | Rethinking Learnable Tree Filter For Generic Feature Transform Highlight: To relax the geometric constraint, we give the analysis by reformulating it as a Markov Random Field and introduce a learnable unary term. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lin Song; Yanwei Li; Zhengkai Jiang; Zeming Li; Xiangyu Zhang; Hongbin Sun; Jian Sun; Nanning Zheng; | code |
| 337 | Self-Supervised Relational Reasoning For Representation Learning Highlight: In this work, we propose a novel self-supervised formulation of relational reasoning that allows a learner to bootstrap a signal from information implicit in unlabeled data. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Massimiliano Patacchiola; Amos J. Storkey; | |
| 338 | Sufficient Dimension Reduction For Classification Using Principal Optimal Transport Direction Highlight: To address this issue, we propose a novel estimation method of sufficient dimension reduction subspace (SDR subspace) using optimal transport. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Cheng Meng; Jun Yu; Jingyi Zhang; Ping Ma; Wenxuan Zhong; | |
| 339 | Fast Epigraphical Projection-based Incremental Algorithms For Wasserstein Distributionally Robust Support Vector Machine Highlight: In this paper, we focus on a family of Wasserstein distributionally robust support vector machine (DRSVM) problems and propose two novel epigraphical projection-based incremental algorithms to solve them. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jiajin Li; Caihua Chen; Anthony Man-Cho So; | |
| 340 | Differentially Private Clustering: Tight Approximation Ratios Highlight: For several basic clustering problems, including Euclidean DensestBall, 1-Cluster, k-means, and k-median, we give efficient differentially private algorithms that achieve essentially the same approximation ratios as those that can be obtained by any non-private algorithm, while incurring only small additive errors. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Badih Ghazi; Ravi Kumar; Pasin Manurangsi; | |
| 341 | On The Power Of Louvain In The Stochastic Block Model Highlight: We provide valuable tools for the analysis of Louvain, but also for many other combinatorial algorithms. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Vincent Cohen-Addad; Adrian Kosowski; Frederik Mallmann-Trenn; David Saulpic; | |
| 342 | Fairness With Overlapping Groups; A Probabilistic Perspective Highlight: In algorithmically fair prediction problems, a standard goal is to ensure the equality of fairness metrics across multiple overlapping groups simultaneously. We reconsider this standard fair classification problem using a probabilistic population analysis, which, in turn, reveals the Bayes-optimal classifier. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Forest Yang; Mouhamadou Cisse; Oluwasanmi O. Koyejo; | |
| 343 | AttendLight: Universal Attention-Based Reinforcement Learning Model For Traffic Signal Control Highlight: We propose AttendLight, an end-to-end Reinforcement Learning (RL) algorithm for the problem of traffic signal control. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Afshin Oroojlooy; Mohammadreza Nazari; Davood Hajinezhad; Jorge Silva; | |
| 344 | Searching For Low-Bit Weights In Quantized Neural Networks Highlight: Thus, we present to regard the discrete weights in an arbitrary quantized neural network as searchable variables, and utilize a differential method to search them accurately. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
zhaohui yang; Yunhe Wang; Kai Han; Chunjing XU; Chao Xu; Dacheng Tao; Chang Xu; | |
| 345 | Adaptive Reduced Rank Regression Highlight: To complement the upper bound, we introduce new techniques for establishing lower bounds on the performance of any algorithm for this problem. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Qiong Wu; Felix MF Wong; Yanhua Li; Zhenming Liu; Varun Kanade; | |
| 346 | From Predictions To Decisions: Using Lookahead Regularization Highlight: For this, we introduce look-ahead regularization which, by anticipating user actions, encourages predictive models to also induce actions that improve outcomes. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Nir Rosenfeld; Anna Hilgard; Sai Srivatsa Ravindranath; David C. Parkes; | |
| 347 | Sequential Bayesian Experimental Design With Variable Cost Structure Highlight: We propose and demonstrate an algorithm that accounts for these variable costs in the refinement decision. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sue Zheng; David Hayden; Jason Pacheco; John W. Fisher III; | |
| 348 | Predictive Inference Is Free With The Jackknife+-after-bootstrap Highlight: In this paper, we propose the jackknife+-after-bootstrap (J+aB), a procedure for constructing a predictive interval, which uses only the available bootstrapped samples and their corresponding fitted models, and is therefore "free" in terms of the cost of model fitting. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Byol Kim; Chen Xu; Rina Foygel Barber; | |
| 349 | Counterfactual Predictions Under Runtime Confounding Highlight: We propose a doubly-robust procedure for learning counterfactual prediction models in this setting. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Amanda Coston; Edward Kennedy; Alexandra Chouldechova; | |
| 350 | Learning Loss For Test-Time Augmentation Highlight: This paper proposes a novel instance-level test- time augmentation that efficiently selects suitable transformations for a test input. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ildoo Kim; Younghoon Kim; Sungwoong Kim; | |
| 351 | Balanced Meta-Softmax For Long-Tailed Visual Recognition Highlight: In this paper, we show that the Softmax function, though used in most classification tasks, gives a biased gradient estimation under the long-tailed setup. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ren Jiawei; Cunjun Yu; shunan sheng; Xiao Ma; Haiyu Zhao; Shuai Yi; hongsheng Li; | |
| 352 | Efficient Exploration Of Reward Functions In Inverse Reinforcement Learning Via Bayesian Optimization Highlight: This paper presents an IRL framework called Bayesian optimization-IRL (BO-IRL) which identifies multiple solutions that are consistent with the expert demonstrations by efficiently exploring the reward function space. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sreejith Balakrishnan; Quoc Phong Nguyen; Bryan Kian Hsiang Low; Harold Soh; | |
| 353 | MDP Homomorphic Networks: Group Symmetries In Reinforcement Learning Highlight: This paper introduces MDP homomorphic networks for deep reinforcement learning. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Elise van der Pol; Daniel Worrall; Herke van Hoof; Frans Oliehoek; Max Welling; | |
| 354 | How Can I Explain This To You? An Empirical Study Of Deep Neural Network Explanation Methods Highlight: We performed a cross-analysis Amazon Mechanical Turk study comparing the popular state-of-the-art explanation methods to empirically determine which are better in explaining model decisions. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jeya Vikranth Jeyakumar; Joseph Noor; Yu-Hsi Cheng; Luis Garcia; Mani Srivastava; | |
| 355 | On The Error Resistance Of Hinge-Loss Minimization Highlight: In this work, we identify a set of conditions on the data under which such surrogate loss minimization algorithms provably learn the correct classifier. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kunal Talwar; | |
| 356 | Munchausen Reinforcement Learning Highlight: Our core contribution stands in a very simple idea: adding the scaled log-policy to the immediate reward. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Nino Vieillard; Olivier Pietquin; Matthieu Geist; | |
| 357 | Object Goal Navigation Using Goal-Oriented Semantic Exploration Highlight: We propose a modular system called, `Goal-Oriented Semantic Exploration’ which builds an episodic semantic map and uses it to explore the environment efficiently based on the goal object category. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Devendra Singh Chaplot; Dhiraj Prakashchand Gandhi; Abhinav Gupta; Russ R. Salakhutdinov; | |
| 358 | Efficient Semidefinite-programming-based Inference For Binary And Multi-class MRFs Highlight: In this paper, we propose an efficient method for computing the partition function or MAP estimate in a pairwise MRF by instead exploiting a recently proposed coordinate-descent-based fast semidefinite solver. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Chirag Pabbaraju; Po-Wei Wang; J. Zico Kolter; | |
| 359 | Funnel-Transformer: Filtering Out Sequential Redundancy For Efficient Language Processing Highlight: With this intuition, we propose Funnel-Transformer which gradually compresses the sequence of hidden states to a shorter one and hence reduces the computation cost. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Zihang Dai; Guokun Lai; Yiming Yang; Quoc Le; | |
| 360 | Semantic Visual Navigation By Watching YouTube Videos Highlight: This paper learns and leverages such semantic cues for navigating to objects of interest in novel environments, by simply watching YouTube videos. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Matthew Chang; Arjun Gupta; Saurabh Gupta; | |
| 361 | Heavy-tailed Representations, Text Polarity Classification & Data Augmentation Highlight: In this paper, we develop a novel method to learn a heavy-tailed embedding with desirable regularity properties regarding the distributional tails, which allows to analyze the points far away from the distribution bulk using the framework of multivariate extreme value theory. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Hamid JALALZAI; Pierre Colombo; Chlo� Clavel; Eric Gaussier; Giovanna Varni; Emmanuel Vignon; Anne Sabourin; | |
| 362 | SuperLoss: A Generic Loss For Robust Curriculum Learning Highlight: We propose instead a simple and generic method that can be applied to a variety of losses and tasks without any change in the learning procedure. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Thibault Castells; Philippe Weinzaepfel; Jerome Revaud; | |
| 363 | CogMol: Target-Specific And Selective Drug Design For COVID-19 Using Deep Generative Models Highlight: In this study, we propose an end-to-end framework, named CogMol (Controlled Generation of Molecules), for designing new drug-like small molecules targeting novel viral proteins with high affinity and off-target selectivity. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Vijil Chenthamarakshan; Payel Das; Samuel Hoffman; Hendrik Strobelt; Inkit Padhi; Kar Wai Lim; Ben Hoover; Matteo Manica; Jannis Born; Teodoro Laino; Aleksandra Mojsilovic; | |
| 364 | Memory Based Trajectory-conditioned Policies For Learning From Sparse Rewards Highlight: In this work, instead of focusing on good experiences with limited diversity, we propose to learn a trajectory-conditioned policy to follow and expand diverse past trajectories from a memory buffer. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yijie Guo; Jongwook Choi; Marcin Moczulski; Shengyu Feng; Samy Bengio; Mohammad Norouzi; Honglak Lee; | |
| 365 | Liberty Or Depth: Deep Bayesian Neural Nets Do Not Need Complex Weight Posterior Approximations Highlight: We challenge the longstanding assumption that the mean-field approximation for variational inference in Bayesian neural networks is severely restrictive, and show this is not the case in deep networks. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sebastian Farquhar; Lewis Smith; Yarin Gal; | |
| 366 | Improving Sample Complexity Bounds For (Natural) Actor-Critic Algorithms Highlight: In contrast, this paper characterizes the convergence rate and sample complexity of AC and NAC under Markovian sampling, with mini-batch data for each iteration, and with actor having general policy class approximation. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tengyu Xu; Zhe Wang; Yingbin Liang; | |
| 367 | Learning Differential Equations That Are Easy To Solve Highlight: We propose a remedy that encourages learned dynamics to be easier to solve. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jacob Kelly; Jesse Bettencourt; Matthew J. Johnson; David K. Duvenaud; | |
| 368 | Stability Of Stochastic Gradient Descent On Nonsmooth Convex Losses Highlight: Specifically, we provide sharp upper and lower bounds for several forms of SGD and full-batch GD on arbitrary Lipschitz nonsmooth convex losses. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Raef Bassily; Vitaly Feldman; Cristobal Guzman; Kunal Talwar; | |
| 369 | Influence-Augmented Online Planning For Complex Environments Highlight: In this work, we propose influence-augmented online planning, a principled method to transform a factored simulator of the entire environment into a local simulator that samples only the state variables that are most relevant to the observation and reward of the planning agent and captures the incoming influence from the rest of the environment using machine learning methods. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jinke He; Miguel Suau de Castro; Frans Oliehoek; | |
| 370 | PAC-Bayes Learning Bounds For Sample-Dependent Priors Highlight: We present a series of new PAC-Bayes learning guarantees for randomized algorithms with sample-dependent priors. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Pranjal Awasthi; Satyen Kale; Stefani Karp; Mehryar Mohri; | |
| 371 | Reward-rational (implicit) Choice: A Unifying Formalism For Reward Learning Highlight: Our key observation is that different types of behavior can be interpreted in a single unifying formalism – as a reward-rational choice that the human is making, often implicitly. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Hong Jun Jeon; Smitha Milli; Anca Dragan; | |
| 372 | Probabilistic Time Series Forecasting With Shape And Temporal Diversity Highlight: In this paper, we address this problem for non-stationary time series, which is very challenging yet crucially important. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Vincent LE GUEN; Nicolas THOME; | |
| 373 | Low Distortion Block-Resampling With Spatially Stochastic Networks Highlight: We formalize and attack the problem of generating new images from old ones that are as diverse as possible, only allowing them to change without restrictions in certain parts of the image while remaining globally consistent. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sarah Hong; Martin Arjovsky; Darryl Barnhart; Ian Thompson; | |
| 374 | Continual Deep Learning By Functional Regularisation Of Memorable Past Highlight: In this paper, we fix this issue by using a new functional-regularisation approach that utilises a few memorable past examples crucial to avoid forgetting. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Pingbo Pan; Siddharth Swaroop; Alexander Immer; Runa Eschenhagen; Richard Turner; Mohammad Emtiyaz E. Khan; | |
| 375 | Distance Encoding: Design Provably More Powerful Neural Networks For Graph Representation Learning Highlight: Here we propose and mathematically analyze a general class of structure-related features, termed Distance Encoding (DE). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Pan Li; Yanbang Wang; Hongwei Wang; Jure Leskovec; | |
| 376 | Fast Fourier Convolution Highlight: In this work, we propose a novel convolutional operator dubbed as fast Fourier convolution (FFC), which has the main hallmarks of non-local receptive fields and cross-scale fusion within the convolutional unit. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Lu Chi; Borui Jiang; Yadong Mu; | |
| 377 | Unsupervised Learning Of Dense Visual Representations Highlight: In this paper, we propose View-Agnostic Dense Representation (VADeR) for unsupervised learning of dense representations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Pedro O. O. Pinheiro; Amjad Almahairi; Ryan Benmalek; Florian Golemo; Aaron C. Courville; | |
| 378 | Higher-Order Certification For Randomized Smoothing Highlight: In this work, we propose a framework to improve the certified safety region for these smoothed classifiers without changing the underlying smoothing scheme. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jeet Mohapatra; Ching-Yun Ko; Tsui-Wei Weng; Pin-Yu Chen; Sijia Liu; Luca Daniel; | |
| 379 | Learning Structured Distributions From Untrusted Batches: Faster And Simpler Highlight: In this paper, we find an appealing way to synthesize the techniques of [JO19] and [CLM19] to give the best of both worlds: an algorithm which runs in polynomial time and can exploit structure in the underlying distribution to achieve sublinear sample complexity. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sitan Chen; Jerry Li; Ankur Moitra; | |
| 380 | Hierarchical Quantized Autoencoders Highlight: This leads us to introduce a novel objective for training hierarchical VQ-VAEs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Will Williams; Sam Ringer; Tom Ash; David MacLeod; Jamie Dougherty; John Hughes; | |
| 381 | Diversity Can Be Transferred: Output Diversification For White- And Black-box Attacks Highlight: To improve the efficiency of these attacks, we propose Output Diversified Sampling (ODS), a novel sampling strategy that attempts to maximize diversity in the target model’s outputs among the generated samples. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yusuke Tashiro; Yang Song; Stefano Ermon; | |
| 382 | POLY-HOOT: Monte-Carlo Planning In Continuous Space MDPs With Non-Asymptotic Analysis Highlight: In this paper, we consider Monte-Carlo planning in an environment with continuous state-action spaces, a much less understood problem with important applications in control and robotics. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Weichao Mao; Kaiqing Zhang; Qiaomin Xie; Tamer Basar; | |
| 383 | AvE: Assistance Via Empowerment Highlight: We propose a new paradigm for assistance by instead increasing the human’s ability to control their environment, and formalize this approach by augmenting reinforcement learning with human empowerment. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yuqing Du; Stas Tiomkin; Emre Kiciman; Daniel Polani; Pieter Abbeel; Anca Dragan; | |
| 384 | Variational Policy Gradient Method For Reinforcement Learning With General Utilities Highlight: In this paper, we consider policy optimization in Markov Decision Problems, where the objective is a general utility function of the state-action occupancy measure, which subsumes several of the aforementioned examples as special cases. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Junyu Zhang; Alec Koppel; Amrit Singh Bedi; Csaba Szepesvari; Mengdi Wang; | |
| 385 | Reverse-engineering Recurrent Neural Network Solutions To A Hierarchical Inference Task For Mice Highlight: We study how recurrent neural networks (RNNs) solve a hierarchical inference task involving two latent variables and disparate timescales separated by 1-2 orders of magnitude. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Rylan Schaeffer; Mikail Khona; Leenoy Meshulam; Brain Laboratory International; Ila Fiete; | |
| 386 | Temporal Positive-unlabeled Learning For Biomedical Hypothesis Generation Via Risk Estimation Highlight: We propose a variational inference model to estimate the positive prior, and incorporate it in the learning of node pair embeddings, which are then used for link prediction. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Uchenna Akujuobi; Jun Chen; Mohamed Elhoseiny; Michael Spranger; Xiangliang Zhang; | |
| 387 | Efficient Low Rank Gaussian Variational Inference For Neural Networks Highlight: By using a new form of the reparametrization trick, we derive a computationally efficient algorithm for performing VI with a Gaussian family with a low-rank plus diagonal covariance structure. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Marcin Tomczak; Siddharth Swaroop; Richard Turner; | |
| 388 | Privacy Amplification Via Random Check-Ins Highlight: In this paper, we focus on conducting iterative methods like DP-SGD in the setting of federated learning (FL) wherein the data is distributed among many devices (clients). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Borja Balle; Peter Kairouz; Brendan McMahan; Om Dipakbhai Thakkar; Abhradeep Thakurta; | |
| 389 | Probabilistic Circuits For Variational Inference In Discrete Graphical Models Highlight: In this paper, we propose a new approach that leverages the tractability of probabilistic circuit models, such as Sum Product Networks (SPN), to compute ELBO gradients exactly (without sampling) for a certain class of densities. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Andy Shih; Stefano Ermon; | |
| 390 | Your Classifier Can Secretly Suffice Multi-Source Domain Adaptation Highlight: In this work, we present a different perspective to MSDA wherein deep models are observed to implicitly align the domains under label supervision. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Naveen Venkat; Jogendra Nath Kundu; Durgesh Singh; Ambareesh Revanur; Venkatesh Babu R; | |
| 391 | Labelling Unlabelled Videos From Scratch With Multi-modal Self-supervision Highlight: In this work, we a) show that unsupervised labelling of a video dataset does not come for free from strong feature encoders and b) propose a novel clustering method that allows pseudo-labelling of a video dataset without any human annotations, by leveraging the natural correspondence between audio and visual modalities. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yuki Asano; Mandela Patrick; Christian Rupprecht; Andrea Vedaldi; | |
| 392 | A Non-Asymptotic Analysis For Stein Variational Gradient Descent Highlight: In this paper, we provide a novel finite time analysis for the SVGD algorithm. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Anna Korba; Adil SALIM; Michael Arbel; Giulia Luise; Arthur Gretton; | |
| 393 | Robust Meta-learning For Mixed Linear Regression With Small Batches Highlight: We introduce a spectral approach that is simultaneously robust under both scenarios. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Weihao Kong; Raghav Somani; Sham Kakade; Sewoong Oh; | |
| 394 | Bayesian Deep Learning And A Probabilistic Perspective Of Generalization Highlight: We show that deep ensembles provide an effective mechanism for approximate Bayesian marginalization, and propose a related approach that further improves the predictive distribution by marginalizing within basins of attraction, without significant overhead. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Andrew Gordon Wilson; Pavel Izmailov; | |
| 395 | Unsupervised Learning Of Object Landmarks Via Self-Training Correspondence Highlight: This paper addresses the problem of unsupervised discovery of object landmarks. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Dimitrios Mallis; Enrique Sanchez; Matthew Bell; Georgios Tzimiropoulos; | code |
| 396 | Randomized Tests For High-dimensional Regression: A More Efficient And Powerful Solution Highlight: In this paper, we answer this question in the affirmative by leveraging the random projection techniques, and propose a testing procedure that blends the classical $F$-test with a random projection step. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yue Li; Ilmun Kim; Yuting Wei; | |
| 397 | Learning Representations From Audio-Visual Spatial Alignment Highlight: We introduce a novel self-supervised pretext task for learning representations from audio-visual content. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Pedro Morgado; Yi Li; Nuno Nvasconcelos; | code |
| 398 | Generative View Synthesis: From Single-view Semantics To Novel-view Images Highlight: We propose to push the envelope further, and introduce Generative View Synthesis (GVS) that can synthesize multiple photorealistic views of a scene given a single semantic map. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tewodros Amberbir Habtegebrial; Varun Jampani; Orazio Gallo; Didier Stricker; | code |
| 399 | Towards More Practical Adversarial Attacks On Graph Neural Networks Highlight: Therefore, we propose a greedy procedure to correct the importance score that takes into account of the diminishing-return pattern. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jiaqi Ma; Shuangrui Ding; Qiaozhu Mei; | |
| 400 | Multi-Task Reinforcement Learning With Soft Modularization Highlight: Thus, instead of naively sharing parameters across tasks, we introduce an explicit modularization technique on policy representation to alleviate this optimization issue. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ruihan Yang; Huazhe Xu; YI WU; Xiaolong Wang; | |
| 401 | Causal Shapley Values: Exploiting Causal Knowledge To Explain Individual Predictions Of Complex Models Highlight: In this paper, we propose a novel framework for computing Shapley values that generalizes recent work that aims to circumvent the independence assumption. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Tom Heskes; Evi Sijben; Ioan Gabriel Bucur; Tom Claassen; | |
| 402 | On The Training Dynamics Of Deep Networks With $L_2$ Regularization Highlight: We study the role of $L_2$ regularization in deep learning, and uncover simple relations between the performance of the model, the $L_2$ coefficient, the learning rate, and the number of training steps. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Aitor Lewkowycz; Guy Gur-Ari; | |
| 403 | Improved Algorithms For Convex-Concave Minimax Optimization Highlight: This paper studies minimax optimization problems minx maxy f(x, y), where f(x, y) is mx-strongly convex with respect to x, my-strongly concave with respect to y and (Lx, Lxy, Ly)-smooth. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yuanhao Wang; Jian Li; | |
| 404 | Deep Variational Instance Segmentation Highlight: In this paper, we propose a novel algorithm that directly utilizes a fully convolutional network (FCN) to predict instance labels. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Jialin Yuan; Chao Chen; Fuxin Li; | |
| 405 | Learning Implicit Functions For Topology-Varying Dense 3D Shape Correspondence Highlight: The goal of this paper is to learn dense 3D shape correspondence for topology-varying objects in an unsupervised manner. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Feng Liu; Xiaoming Liu; | |
| 406 | Deep Multimodal Fusion By Channel Exchanging Highlight: To this end, this paper proposes Channel-Exchanging-Network (CEN), a parameter-free multimodal fusion framework that dynamically exchanges channels between sub-networks of different modalities. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yikai Wang; Wenbing Huang; Fuchun Sun; Tingyang Xu; Yu Rong; Junzhou Huang; | |
| 407 | Hierarchically Organized Latent Modules For Exploratory Search In Morphogenetic Systems Highlight: In this paper, we motivate the need for what we call Meta-diversity search, arguing that there is not a unique ground truth interesting diversity as it strongly depends on the final observer and its motives. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Mayalen Etcheverry; Cl�ment Moulin-Frier; Pierre-Yves Oudeyer; | |
| 408 | AI Feynman 2.0: Pareto-optimal Symbolic Regression Exploiting Graph Modularity Highlight: We present an improved method for symbolic regression that seeks to fit data to formulas that are Pareto-optimal, in the sense of having the best accuracy for a given complexity. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Silviu-Marian Udrescu; Andrew Tan; Jiahai Feng; Orisvaldo Neto; Tailin Wu; Max Tegmark; | |
| 409 | Delay And Cooperation In Nonstochastic Linear Bandits Highlight: This paper offers a nearly optimal algorithm for online linear optimization with delayed bandit feedback. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shinji Ito; Daisuke Hatano; Hanna Sumita; Kei Takemura; Takuro Fukunaga; Naonori Kakimura; Ken-Ichi Kawarabayashi; | |
| 410 | Probabilistic Orientation Estimation With Matrix Fisher Distributions Highlight: This paper focuses on estimating probability distributions over the set of 3D ro- tations (SO(3)) using deep neural networks. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
David Mohlin; Josephine Sullivan; G�rald Bianchi; | code |
| 411 | Minimax Dynamics Of Optimally Balanced Spiking Networks Of Excitatory And Inhibitory Neurons Highlight: Overall, we present a novel normative modeling approach for spiking E-I networks, going beyond the widely-used energy-minimizing networks that violate Dale’s law. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Qianyi Li; Cengiz Pehlevan; | |
| 412 | Telescoping Density-Ratio Estimation Highlight: To resolve this limitation, we introduce a new framework, telescoping density-ratio estimation (TRE), that enables the estimation of ratios between highly dissimilar densities in high-dimensional spaces. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Benjamin Rhodes; Kai Xu; Michael U. Gutmann; | |
| 413 | Towards Deeper Graph Neural Networks With Differentiable Group Normalization Highlight: To bridge the gap, we introduce two over-smoothing metrics and a novel technique, i.e., differentiable group normalization (DGN). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kaixiong Zhou; Xiao Huang; Yuening Li; Daochen Zha; Rui Chen; Xia Hu; | |
| 414 | Stochastic Optimization For Performative Prediction Highlight: We initiate the study of stochastic optimization for performative prediction. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Celestine Mendler-D�nner; Juan Perdomo; Tijana Zrnic; Moritz Hardt; | |
| 415 | Learning Differentiable Programs With Admissible Neural Heuristics Highlight: We study the problem of learning differentiable functions expressed as programs in a domain-specific language. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Ameesh Shah; Eric Zhan; Jennifer Sun; Abhinav Verma; Yisong Yue; Swarat Chaudhuri; | |
| 416 | Improved Guarantees And A Multiple-descent Curve For Column Subset Selection And The Nystrom Method Highlight: We develop techniques which exploit spectral properties of the data matrix to obtain improved approximation guarantees which go beyond the standard worst-case analysis. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Michal Derezinski; Rajiv Khanna; Michael W. Mahoney; | |
| 417 | Domain Adaptation As A Problem Of Inference On Graphical Models Highlight: To develop an automated way of domain adaptation with multiple source domains, we propose to use a graphical model as a compact way to encode the change property of the joint distribution, which can be learned from data, and then view domain adaptation as a problem of Bayesian inference on the graphical models. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kun Zhang; Mingming Gong; Petar Stojanov; Biwei Huang; QINGSONG LIU; Clark Glymour; | |
| 418 | Network Size And Size Of The Weights In Memorization With Two-layers Neural Networks Highlight: In contrast we propose a new training procedure for ReLU networks, based on {\em complex} (as opposed to {\em real}) recombination of the neurons, for which we show approximate memorization with both $O\left(\frac{n}{d} \cdot \frac{\log(1/\epsilon)}{\epsilon}\right)$ neurons, as well as nearly-optimal size of the weights. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Sebastien Bubeck; Ronen Eldan; Yin Tat Lee; Dan Mikulincer; | |
| 419 | Certifying Strategyproof Auction Networks Highlight: We propose ways to explicitly verify strategyproofness under a particular valuation profile using techniques from the neural network verification literature. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Michael Curry; Ping-yeh Chiang; Tom Goldstein; John Dickerson; | |
| 420 | Continual Learning Of Control Primitives : Skill Discovery Via Reset-Games Highlight: In this work, we show how a single method can allow an agent to acquire skills with minimal supervision while removing the need for resets. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Kelvin Xu; Siddharth Verma; Chelsea Finn; Sergey Levine; | |
| 421 | HOI Analysis: Integrating And Decomposing Human-Object Interaction Highlight: In analogy to Harmonic Analysis, whose goal is to study how to represent the signals with the superposition of basic waves, we propose the HOI Analysis. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Yong-Lu Li; Xinpeng Liu; Xiaoqian Wu; Yizhuo Li; Cewu Lu; | code |
| 422 | Strongly Local P-norm-cut Algorithms For Semi-supervised Learning And Local Graph Clustering Highlight: In this paper, we propose a generalization of the objective function behind these methods involving p-norms. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Meng Liu; David F. Gleich; | |
| 423 | Deep Direct Likelihood Knockoffs Highlight: We develop Deep Direct Likelihood Knockoffs (DDLK), which directly minimizes the KL divergence implied by the knockoff swap property. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Mukund Sudarshan; Wesley Tansey; Rajesh Ranganath; | |
| 424 | Meta-Neighborhoods Highlight: In this work, we step forward in this direction and propose a semi-parametric method, Meta-Neighborhoods, where predictions are made adaptively to the neighborhood of the input. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Siyuan Shan; Yang Li; Junier B. Oliva; | |
| 425 | Neural Dynamic Policies For End-to-End Sensorimotor Learning Highlight: In this work, we begin to close this gap and embed dynamics structure into deep neural network-based policies by reparameterizing action spaces with differential equations. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Shikhar Bahl; Mustafa Mukadam; Abhinav Gupta; Deepak Pathak; | code |
| 426 | A New Inference Approach For Training Shallow And Deep Generalized Linear Models Of Noisy Interacting Neurons Highlight: Here, we develop a two-step inference strategy that allows us to train robust generalised linear models of interacting neurons, by explicitly separating the effects of correlations in the stimulus from network interactions in each training step. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Gabriel Mahuas; Giulio Isacchini; Olivier Marre; Ulisse Ferrari; Thierry Mora; | |
| 427 | Decision-Making With Auto-Encoding Variational Bayes Highlight: Motivated by these theoretical results, we propose learning several approximate proposals for the best model and combining them using multiple importance sampling for decision-making. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Romain Lopez; Pierre Boyeau; Nir Yosef; Michael Jordan; Jeffrey Regier; | |
| 428 | Attribution Preservation In Network Compression For Reliable Network Interpretation Highlight: In this paper, we show that these seemingly unrelated techniques conflict with each other as network compression deforms the produced attributions, which could lead to dire consequences for mission-critical applications. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Geondo Park; June Yong Yang; Sung Ju Hwang; Eunho Yang; | |
| 429 | Feature Importance Ranking For Deep Learning Highlight: In this paper, we propose a novel dual-net architecture consisting of operator and selector for discovery of an optimal feature subset of a fixed size and ranking the importance of those features in the optimal subset simultaneously. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Maksymilian Wojtas; Ke Chen; | code |
| 430 | Causal Estimation With Functional Confounders Highlight: We study causal inference when the true confounder value can be expressed as a function of the observed data; we call this setting estimation with functional confounders (EFC). Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Aahlad Manas Puli; Adler Perotte ; Rajesh Ranganath; | |
| 431 | Model Inversion Networks For Model-Based Optimization Highlight: We propose to address such problems with model inversion networks (MINs), which learn an inverse mapping from scores to inputs. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Aviral Kumar; Sergey Levine; | |
| 432 | Hausdorff Dimension, Heavy Tails, And Generalization In Neural Networks Highlight: Aiming to bridge this gap, in this paper, we prove generalization bounds for SGD under the assumption that its trajectories can be well-approximated by a \emph{Feller process}, which defines a rich class of Markov processes that include several recent SDE representations (both Brownian or heavy-tailed) as its special case. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Umut Simsekli; Ozan Sener; George Deligiannidis; Murat A. Erdogdu; | |
| 433 | Exact Expressions For Double Descent And Implicit Regularization Via Surrogate Random Design Highlight: We provide the first exact non-asymptotic expressions for double descent of the minimum norm linear estimator. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Michal Derezinski; Feynman T. Liang; Michael W. Mahoney; | |
| 434 | Certifying Confidence Via Randomized Smoothing Highlight: In this work, we propose a method to generate certified radii for the prediction confidence of the smoothed classifier. Related Papers Related Patents Related Grants Related Orgs Related Experts Details |
Aounon Kumar; Alexander Levine; Soheil Feizi; Tom Goldstein; | code |
| 435 | Learning Physical Constraints With Neural Projections Highlight: We propose a new family of neural networks to predict the behaviors of physical systems by learning their underpinning constraints. Related Papers |