Paper Digest: COLT 2026 Papers & Highlights
The Annual Conference on Learning Theory (COLT) is one of the premier international conferences on learning theory. To help the community quickly catch up on the work presented in this conference, Paper Digest Team processed all accepted papers, and generated one highlight sentence (typically the main topic) for each paper. Readers are encouraged to read these machine generated highlights to quickly get the main idea of each paper.
To search for papers presented at COLT-2026 on a specific topic, please make use of the search by venue (COLT-2026) service. To summarize the latest research published at COLT-2026 on a specific topic, you can utilize the review by venue (COLT-2026) service. If you are interested in browsing papers by author, we have a comprehensive list of ~ 500 authors (COLT-2026).
Since 2018, Paper Digest has built a foundation of data spanning decades of conferences, journals, and research topics. The platform features a daily digest service that sifts through tens of thousands of new papers, clinical trials, news articles, and community posts, filtering the noise to highlight what matters most to specific interests. Beyond daily updates, dozens of built-in research tools streamline the academic workflow, supporting efficient reading and writing, comprehensive literature reviews, and automated research report generation.
Paper Digest Team
New York City, New York, 10017
team@paperdigest.org
TABLE 1: Paper Digest: COLT 2026 Papers & Highlights
| Paper | Author(s) | |
|---|---|---|
| 1 | Information-computation Gaps in Quantum Learning Via Low-degree Likelihood Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we extend this framework to the quantum setting and show a number of new information-computation gaps for quantum learning. |
Sitan Chen; Weiyuan Gong; Jonas Haferkamp; Yihui Quek; |
| 2 | Provable Learning of Random Hierarchy Models and Hierarchical Shallow-to-Deep Chaining Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we consider Random Hierarchy Models — a hierarchical context-free grammar introduced by Cagnetta et al. (2024) and conjectured to separate deep and shallow networks. |
Yunwei Ren; Yatin Dandi; Florent Krzakala; Jason D. Lee; |
| 3 | Query Efficient Structured Matrix Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the problem of learning a structured approximation (low-rank, sparse, banded, etc.) to an unknown matrix $\boldsymbol{\mathbf{A}}$ given access to matrix-vector product (matvec) queries of the form $\boldsymbol{\mathbf{x}} \mapsto \boldsymbol{\mathbf{A}}\boldsymbol{\mathbf{x}}$ and $\boldsymbol{\mathbf{x}} \mapsto \boldsymbol{\mathbf{A}}^\transpose \boldsymbol{\mathbf{x}}$. |
Noah Amsel; Pratyush Avi; Tyler Chen; Feyza Duman Keles; Chinmay Hegde; Christopher Musco; Cameron Musco; David Persson; |
| 4 | Optimal Inference Schedules for Masked Diffusion Models Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we give a new, \emph{exact} characterization of the expected divergence between the true distribution and the sampled distribution, for any distribution and any unmasking schedule for the sampler, showing an elegant connection to the theory of \emph{univariate function approximation}. |
Sitan Chen; Kevin Cong; Jerry Li; |
| 5 | How Many Features Can A Language Model Store Under The Linear Representation Hypothesis? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce a mathematical framework for the linear representation hypothesis (LRH), which asserts that intermediate layers of language models store features linearly. |
Nikhil Garg; Jon Kleinberg; Kenny Peng; |
| 6 | Can SGD Select Good Fishermen? Local Convergence Under Self-Selection Biases (Extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We revisit the problem of estimating $k$ linear regressors in $d$ dimensions from samples affected by self-selection bias under the maximum selection rule. |
Alkis Kalavasis; Anay Mehrotra; Felix Zhou; |
| 7 | Graph Neural Networks Extrapolate Out-of-distribution for Shortest Paths Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: One promising approach for achieving robust OOD generalization is the framework of neural algorithmic alignment, which incorporates ideas from classical algorithms by designing neural architectures that resemble specific algorithmic paradigms (e.g. dynamic programming). |
Robert R. Nerem; Samantha Chen; Sanjoy Dasgupta; Yusu Wang; |
| 8 | On The Statistical Query Complexity of Learning Semiautomata: A Random Walk Approach Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: By applying tools from Fourier analysis and the representation theory of the symmetric group, we obtain tight spectral gap bounds, demonstrating that after a polynomial number of steps in the number of states, distinct semiautomata become nearly uncorrelated, yielding the desired hardness result. |
George Giapitzakis; Kimon Fountoulakis; Eshaan Nichani; Jason D. Lee; |
| 9 | Invited Open Problem: Is The Power of Deep Learning Over Linear Models Inherently Distribution Dependent? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: We ask whether distribution-independent SQ learning implies low dimension complexity, and whether anything learnable with (S)GD on a (benign) neural network under any input … |
Vitaly Feldman; Pritish Kamath; Nathan Srebro; |
| 10 | When Both Layers Learn: Training Dynamics of Representing Linear Models Via ReLU Networks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we study the gradient descent dynamics for jointly training both layers of a one-hidden-layer ReLU network to fit a linear target function. |
Berk Tinaz; Changzhi Xie; Mahdi Soltanolkotabi; |
| 11 | Learning to Reason with Curriculum I: Provable Benefits of Autocurriculum Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Are these costs fundamental, or can they be reduced through better algorithmic design? We show that \textit{autocurriculum}—where the model uses its own performance to decide which problems to focus training on—provably improves upon standard training recipes for both supervised fine-tuning (SFT) and reinforcement learning (RL). |
Nived Rajaraman; Audrey Huang; Miro Dudik; Rob Schapire; Dylan Foster; Akshay Krishnamurthy; |
| 12 | Steering Diffusion Models with Quadratic Rewards: A Fine-grained Analysis Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we consider the task of sampling from a reward-tilted diffusion model—that is, sampling from $p^{\star}(x) \propto p(x) \exp(r(x))$—given a reward function $r$ and pre-trained diffusion oracle for $p$. |
Ankur Moitra; Andrej Risteski; Dhruv Rohatgi; |
| 13 | Convergence of Continual Learning in Homogeneous Deep Networks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We characterize weakly regularized continual classification in homogeneous models as sequential projections onto task margin sets. |
Matan Schliserman; Gon Buzaglo; Itay Evron; Daniel Soudry; |
| 14 | Invited Open Problem: Online Optimization of Piecewise-Lipschitz Functions with Applications to Data-Driven Algorithm Design Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Classical online optimization theory focuses on regret guarantees for convex Lipschitz functions. |
Maria-Florina Balcan; Wesley Pegden; Dravyansh Sharma; |
| 15 | Tight Sample Complexity of Transformers Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We tightly characterize the VC dimension of depth-$L$ Transformers with a total of $W$ parameters, mapping an input sequence of length $T$ to a single output, establishing an upper bound of $O(L W \log (T W))$ and a nearly matching lower bound of $\Omega(L W \log (T W / L))$. |
Chenxiao Yang; Nathan Srebro; Zhiyuan Li; |
| 16 | CONVERGENCE RATES FOR DISTRIBUTION MATCHING WITH SLICED OPTIMAL TRANSPORT Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the slice-matching scheme, an efficient iterative method for distribution matching based on sliced optimal transport. |
Gauthier Thurin; Claire Boyer; Kimia Nadjahi; |
| 17 | Estimating Ising Models in Total Variation Distance Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our main contribution is a unified analysis of the Maximum Pseudo-Likelihood Estimator (MPLE) for two general classes of Ising models. |
Constantinos Daskalakis; Vardis Kandiros; Rui Yao; |
| 18 | Optimal Learning Rate Schedules Under Functional Scaling Laws: Power Decay and Warmup–Stable–Decay (Extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We next study the practical setting where the decay shape is fixed and only the peak LR is tuned. To separate these two design choices, we introduce a family of fractional LR schedules that decouple peak-LR tuning from decay-shape design. |
Binghui Li; Zilin Wang; Fengling Chen; Shiyang Zhao; Ruiheng Zheng; Lei Wu; |
| 19 | Instance-optimal High-precision Shadow Tomography with Few-copy Measurements: A Metrological Approach Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: More concretely, we study the problem of learning expectation values of a given set of observables of an unknown quantum state to precision $\epsilon$ in $L_p$-norm, using (possibly adaptive) measurements that act on one or a few copies at a time, and we are interested in the regime that $\epsilon$ is below some concrete and potentially dimension-dependent threshold. |
Senrui Chen; Weiyuan Gong; Sisi Zhou; |
| 20 | Swap Regret Minimization Through Response-Based Approachability Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we develop a significantly simpler, computationally efficient algorithm that guarantees $O(d \sqrt{T})$ linear swap regret for a general convex set that has been preconditioned via the John ellipsoid. |
Ioannis Anagnostides; Gabriele Farina; Maxwell Fishelson; Haipeng Luo; Jon Schneider; |
| 21 | Toward Simultaneously Optimal Regret in U-Calibration Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Existing U-calibration algorithms achieve worst-case optimal $O(\sqrt{T})$ regret for every bounded proper loss, but they fail to adapt to easier losses: as we show, even for smooth losses such as squared loss, they incur $\Omega(\sqrt{T})$ regret instead of the optimal $O(\log T)$ regret. In this work, we show that this limitation is not inherent. |
Rafael Frongillo; Haipeng Luo; Nishant A. Mehta; Jon Schneider; |
| 22 | Diffusion-Network Alignment: An Efficient Algorithm and Explicit Probability Bounds Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Different from the classic network alignment where both networks are fully observed, this model captures the information asymmetry of two networks. To solve this problem, this paper presents an efficient algorithm based on tree correlation tests to extract alignment information from local neighborhoods. |
Ziao Wang; Lei Ying; |
| 23 | Risk Comparisons in Linear Regression: Implicit Regularization Dominates Explicit Regularization (Extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Moving beyond minimax theory, this work provides \emph{instance-wise} comparisons of the finite-sample risks for these algorithms on any well-specified linear regression problem.While it is known that for certain problems GD can be polynomially better than SGD, the reverse is also true: we construct problems, inspired by \emph{benign overfitting} theory, where optimally stopped GD is polynomially worse. |
Jingfeng Wu; Peter L. Bartlett; Sham M. Kakade; Jason D. Lee; Bin Yu; |
| 24 | Optimism Stabilizes Thompson Sampling for Adaptive Inference Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We identify optimism as a general mechanism for stabilizing Thompson sampling. |
Shunxing Yan; Han Zhong; |
| 25 | Efficient Sampling with Discrete Diffusion Models: Sharp and Adaptive Guarantees Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we study the sampling efficiency of score-based discrete diffusion models under a continuous-time Markov chain (CTMC) formulation, with a focus on $\tau$-leaping-based samplers. |
Daniil Dmitriev; Zhihan Huang; Yuting Wei; |
| 26 | Rigorous Asymptotics for First-Order Algorithms Through The Dynamical Cavity Method Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we formalize the dynamical cavity method and use it to give a new proof of the DMFT equations for General First Order Methods, a broad class of dynamics encompassing algorithms such as Gradient Descent and Approximate Message Passing. |
Yatin Dandi; David Gamarnik; Francisco Pernice; Lenka Zdeborová; |
| 27 | On The Asymptotics of Self-Supervised Pre-training: Two-Stage M-Estimation and Representation Symmetry Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While a growing body of work has begun to analyze this paradigm, existing bounds leave open the question of how sharp current rates are, and whether they accurately capture the complex interaction between pre-training and fine-tuning. In this paper, we address this gap by developing an asymptotic theory of pre-training via two-stage $M$-estimation. |
Mohammad Tinati; Stephen Tu; |
| 28 | Sharp Analysis of Linear Ensemble Sampling Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We analyse linear ensemble sampling (ES) with standard Gaussian perturbations in stochastic linear bandits. |
David Janz; Arya Akhavan; Csaba Szepesvári; |
| 29 | Is Memorization Helpful or Harmful? Prior Information Sets The Threshold Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We examine the connection between training error and generalization error for arbitrary estimating procedures, working in an overparameterized linear model under general priors in a Bayesian setup. |
Chen Cheng; Rina Foygel Barber; |
| 30 | DDPM Score Matching and Distribution Learning (Extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present a framework reducing the other two forms of distribution learning to score estimation, which has various implications in statistical and computational learning theory: parameter estimation, where denoising score matching in DDPMs is asymptotically efficient; density estimation, where estimated scores can be lifted to a $(\epsilon,\delta)$-PAC density estimator and yield minimax rates over Hölder classes and a quasi-polynomial PAC density estimation algorithm for Gaussian location mixtures; and lower bounds for score estimation, where PAC density estimation yields computational lower bounds for score estimation of general distribution families and cryptographic lower bounds for score estimation of general Gaussian mixture models. |
Sinho Chewi; Alkis Kalavasis; Anay Mehrotra; Omar Montasser; |
| 31 | Adaptive Matrix Online Learning Through Smoothing with Guarantees for Nonsmooth Nonconvex Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study online linear optimization with matrix variables constrained by the operator norm, a setting where the geometry renders designing data-dependent and efficient adaptive algorithms challenging. |
Ruichen Jiang; Zakaria Mhammedi; Mehryar Mohri; Aryan Mokhtari; |
| 32 | Almost Linear Convergence Under Minimal Score Assumptions: Quantized Transition Diffusion Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we develop an improved generative modeling method by introducing Quantized Transition Diffusion (QTD), a framework that reformulates continuous diffusion into a discrete generation problem through spatial quantization and the parameterization of zeroth-order information (e.g., density ratios). |
Xunpeng Huang; Yingyu Lin; Lijing Kuang; Hanze Dong; Difan Zou; Yian Ma; Tong Zhang; |
| 33 | Language Identification with Succinct Machine-Independent Traces Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This recent work has shown positive results for language identification in the presence of such computational traces, but the traces in these positive results come from explicit automata-theoretic machine models that generate the language, where the underlying vocabulary of tokens for the traces is very large. In this paper, we address two fundamental issues left open by this line of work: can we achieve positive results with traces that use only a small alphabet, and can we define traces directly from the language itself, without requiring an underlying machine model that generates it? |
Moses Charikar; Jon Kleinberg; Chirag Pabbaraju; |
| 34 | Self-Normalized Martingales and Uniform Regret Bounds for Linear Regression Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We characterize when scale-invariant upper bounds on self-normalized martingales are possible. |
Fan Chen; Jian Qian; Alexander Rakhlin; Nikita Zhivotovskiy; |
| 35 | Faster Newton Methods for Convex and Nonconvex Optimization in Gradient Complexity Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose new methods that achieve an improved complexity of $\mathcal{O}( \bar d + \bar d^{1/3} \epsilon^{-3/2})$ and $\mathcal{O}( (\bar d + \bar d^{13/21} \epsilon^{-2/7}) \ln \bar d)$ for nonconvex and convex optimization, respectively, improving best-known results for both setups. |
Lesi Chen; Chengchang Liu; Luo Luo; Jingzhao Zhang; |
| 36 | Efficient Learning and Symmetry Discovery Under Exact Invariances Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we resolve both challenges. |
Ashkan Soleymani; Behrooz Tahmasebi; Patrick Jaillet; Stefanie Jegelka; |
| 37 | Universal Priors: Solving Empirical Bayes Via Bayesian Inference and Pretraining Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We theoretically justify the recent empirical finding of Teh et al. (2025) that a transformer pretrained on synthetically generated data achieves strong performance on empirical Bayes (EB) problems. We take an indirect approach to this question: rather than analyzing the model architecture or training dynamics, we ask why a pretrained Bayes estimator, trained under a prespecified training distribution, can adapt to arbitrary test distributions. |
Nick Cannella; Anzo Teh; Yanjun Han; Yury Polyanskiy; |
| 38 | Space-Efficient Language Generation in The Limit Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In particular, we present a streaming algorithm using $\mathrm{poly}(s,k)$ space that converges to a hypothesis with generation gap $\Delta = O(k^{2s-2})$. |
Nicolas Flammarion; Chirag Pabbaraju; Hristo Papazov; Miltiadis Stouras; Ola Svensson; |
| 39 | High-Accuracy Log-Concave Sampling with Stochastic Queries Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We show that high-accuracy guarantees for log-concave sampling—that is, iteration and query complexities which scale as $\mathrm{poly}\log(1/\delta)$, where $\delta$ is the desired target accuracy—are achievable using stochastic gradients with sub-exponential tails. |
Fan Chen; Sinho Chewi; Constantinos Daskalakis; Alexander Rakhlin; |
| 40 | Tight List Replicability Bounds Via A Novel Sphere Covering Theorem Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In recent years, list replicability has emerged as a framework for formalizing reproducibility in learning theory. |
Ari Blondal; Hamed Hatami; Pooya Hatami; Chavdar Lalov; Sivan Tretiak; |
| 41 | Gradient-Variation Regret Bounds for Unconstrained Online Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We develop parameter-free algorithms for unconstrained online learning with regret guarantees that scale with the gradient variation $V_T(u) = \sum_{t=2}^T \|\nabla f_t(u)-\nabla f_{t-1}(u)\|^2$. |
Yuheng Zhao; Andrew Jacobsen; Nicolò Cesa-Bianchi; Peng Zhao; |
| 42 | Functional Stochastic Localization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Motivated by sampling under non-Euclidean geometries and the mirror descent algorithm in optimization, we develop a functional generalization of Eldan’s process that replaces Gaussian regularization with regularization by any positive integer multiple of a log-Laplace transform. |
Anming Gu; Bobby Shi; Kevin Tian; |
| 43 | Calibeating Made Simple Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study calibeating, the problem of post-processing external forecasts online to minimize cumulative losses and match an informativeness-based benchmark. |
Yurong Chen; Zhiyi Huang; Michael I. Jordan; Haipeng Luo; |
| 44 | Language Generation with Infinite Contamination Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: A recent line of work studies language generation in the limit, a formal model of language learning where an algorithm observes an adversarially generated enumeration of strings from an unknown target language $K$ and must eventually generate new, unseen strings from $K$. |
Anay Mehrotra; Grigoris Velegkas; Xifan Yu; Felix Zhou; |
| 45 | Differentially Private Language Generation and Identification in The Limit (Extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We consider the \emph{continual release} model, where a generator must eventually output a stream of valid strings while protecting the privacy of the entire input sequence. |
Anay Mehrotra; Grigoris Velegkas; Xifan Yu; Felix Zhou; |
| 46 | Learning Conditional Averages Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce the problem of learning \emph{conditional averages} in the PAC framework. |
Marco Bressan; Nataly Brukhim; Nicolò Cesa-Bianchi; Emmanuel Esposito; Yishay Mansour; Shay Moran; Maximilian Thiessen; |
| 47 | Limitations of SGD for Multi-Index Models Beyond Statistical Queries Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Moreover, many analyses of SGD for challenging problems rely on non-trivial algorithmic modifications, such as restricting the SGD trajectory to the sphere or using very small learning rates. To address these shortcomings, we develop a new, non-SQ framework to study the limitations of standard vanilla SGD, for single-index and multi-index models (namely, when the target function depends on a low-dimensional projection of the inputs). |
Daniel Barzilai; Ohad Shamir; |
| 48 | Model Agreement Via Anchoring Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We develop a simple general technique for proving bounds on independent model disagreement based on \emph{anchoring} to the average of two models within the analysis. |
Eric Eaton; Surbhi Goel; Marcel Hussing; Michael Kearns; Aaron Roth; Sikata Bela Sengupta; Jessica Sorrell; |
| 49 | Recovery of Planted Subgraphs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we study the exact recovery of an arbitrary planted subgraph $\Gamma = \Gamma_n$ embedded in a dense Erdős–Rényi random graph $\mathcal{G}(n,q_n)$, where edges within $\Gamma$ are present independently with probability $p_n > q_n$. |
Wasim Huleihel; |
| 50 | Near-optimal Swap Regret Minimization for Convex Losses Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We give a randomized online algorithm that guarantees near-optimal $\widetilde{O}(\sqrt{T})$ expected swap regret against any sequence of $T$ adaptively chosen Lipschitz convex losses on the unit interval. |
Lunjia Hu; Jon Schneider; Yifan Wu; |
| 51 | On The Complexity of Best-Arm Identification in Non-Stationary Linear Bandits Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Motivated by the ideas underlying our lower bound, we propose the \textit{Adjacent-optimal design}, a specialization of the well-known $\mathcal{XY}$-optimal design, and develop the \textsf{Adjacent-BAI} algorithm. |
Leo Maynard-Zhang; Zhihan Xiong; Kevin Jamieson; Maryam Fazel; |
| 52 | Near-Optimal Regret for Distributed Adversarial Bandits: A Black-Box Approach Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study distributed adversarial bandits, where $N$ agents cooperate to minimize the global average loss while observing only their own local losses. |
Hao Qiu; Mengxiao Zhang; Nicolò Cesa-Bianchi; |
| 53 | Reconstructing Riemannian Metrics From Random Geometric Graphs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work we consider a similar and arguably more natural problem where the metric is the Riemannian metric on the manifold. |
Han Huang; Pakawut Jiradilok; Elchanan Mossel; |
| 54 | Regret Minimization with Adaptive Opponents in Repeated Games Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we study regret minimization in repeated games with \emph{adaptive} opponents whose strategies may depend on the histories of play. |
Mingyang Liu; Asuman Ozdaglar; Tiancheng Yu; Kaiqing Zhang; |
| 55 | A Characterization of List Language Identification in The Limit Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the problem of language identification in the limit, where given a sequence of examples from a target language, the goal of the learner is to output a sequence of guesses for the target language such that all the guesses beyond some finite time are correct. |
Moses Charikar; Chirag Pabbaraju; Ambuj Tewari; |
| 56 | Random Reshuffling Dominates Stochastic Gradient Descent Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, for the first time, we prove that {\textsf{RR}} dominates {\textsf{SGD}} in smooth convex optimization under any reasonable stepsize after any finite number of epochs, thereby addressing a longstanding open question. |
Zijian Liu; |
| 57 | Optimal Hardness of Online Algorithms for Large Common Induced Subgraphs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the problem of efficiently finding large common induced subgraphs of two independent Erdős–Rényi random graphs $G_1, G_2 \sim \mathbb{G}(n,1/2)$. |
David Gamarnik; Miklós Z. Rácz; Gabe Schoenbach; |
| 58 | Adversarial Learning in Games with Bandit Feedback: Logarithmic Pure-Strategy Maximin Regret Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Crucially, we prove an information-theoretic lower bound showing that the dependence on $c$ is necessary. To overcome this hardness, we turn to the informed setting and introduce Maximin-UCB, which obtains another regret bound of the form $\mathcal{O}(c’ \log T)$ for a different game-dependent parameter $c’$ that could potentially be much smaller than $c$. |
Shinji Ito; Haipeng Luo; Arnab Maiti; Taira Tsuchiya; Yue Wu; |
| 59 | Partition Function Estimation Under Bounded $f$-Divergence Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Along the way we introduce new technical tools including new connections between coverage and $f$-divergences as well as a generalization of the classical Paley-Zygmund inequality. |
Adam Block; Abhishek Shetty; |
| 60 | Almost Sure Null Bankruptcy of Testing-by-betting Strategies Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Under the null, a strategy induces a wealth martingale converging to some random variable that can be zero (bankrupt) or non-zero (non-bankrupt, e.g. when it eventually stops betting). In this paper, we show the conceptually intuitive but technically nontrivial fact that these strategies (universal portfolio, Krichevsky-Trofimov, GRAPA, hedging, etc.) all go bankrupt with probability one, under any non-degenerate null distribution. |
Hongjian Wang; Shubhada Agrawal; Aaditya Ramdas; |
| 61 | The Sample Complexity of Multiclass and Sparse Contextual Bandits Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We further extend this bound to general Natarajan classes and complement it with a matching lower bound (up to logarithmic factors), thereby closing a substantial gap left by prior work (Erez et al., 2024a,b; Erez and Koren, 2025), which incurred an additional $\Theta(|\mathcal{A}|^9)$ dependence. We obtain these results via two complementary approaches. |
Liad Erez; Fan Chen; Alon Cohen; Tomer Koren; Yishay Mansour; Shay Moran; Alexander Rakhlin; |
| 62 | Unified Framework of Distributional Regret in Multi-Armed Bandits and Reinforcement Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present a simple UCBVI-style algorithm with exploration bonus $\min{c_{1,k}/N, c_{2,k}/\sqrt{N}}$, where $N$ denotes the visit count and $(c_{1,k},c_{2,k})$ are user-specified parameters. |
Harin Lee; Min-hwan Oh; |
| 63 | An Empirical Bayes Perspective on Heteroskedastic Mean Estimation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While, with known variances, a simple linear estimator attains the smallest mean squared error, estimation without this knowledge is challenging due to the large number of nuisance parameters. We propose a simple and principled approach based on empirical Bayes: model the observations as if they were i.i.d. from a normal scale mixture and compute the profile maximum likelihood estimator (MLE) for the mean, treating the nonparametric mixing distribution as nuisance. |
Yanjun Han; Abhishek Shetty; Jacob Shkrob; |
| 64 | Ripple Mechanisms for Discrete and Private Statistics Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study \emph{ripple mechanisms} for pure differentially private computation of discrete statistics. |
Matthew Joseph; Alex Kulesza; Yuyan Wang; Alexander Yu; |
| 65 | Learning Depth-3 Circuits Via Quantum Agnostic Boosting Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: We initiate the study of quantum agnostic learning of phase states with respect to a function class $C \subseteq {c:{0,1}^n\rightarrow {0,1}}$: given copies of an unknown … |
Srinivasan Arunachalam; Arkopal Dutt; Alexandru Gheorghiu; Michael De Oliveira; |
| 66 | Separating Oblivious and Adaptive Models of Variable Selection (Extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we investigate the statistical and computational landscapes of sparse recovery with $\ell_\infty$ error guarantees. |
Ziyun Chen; Jerry Li; Kevin Tian; Yusong Zhu; |
| 67 | Compact Geometric Representations of Hierarchies Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we investigate compact reachability embeddings for more general graph classes and provide theoretical guarantees for representing hierarchies using embeddings whose dimension depends on structural graph parameters. |
Prashant Gokhale; Piotr Indyk; Yuhao Liu; Sandeep Silwal; Tony Wang; Haike Xu; |
| 68 | On The Curse of Dimensionality in Private Sparse Covariance Estimation and PCA Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study high-dimensional differentially private (DP) covariance estimation in the operator norm, and principal component analysis (PCA), under $k$-row-column sparsity ($k$-RCS) of the covariance matrix. |
Syamantak Kumar; Shourya Pandey; Purnamrita Sarkar; Kevin Tian; |
| 69 | Fast Score-Based Sampling Via Log-Concave Reductions Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We show how, in some generality, the availability of scores allows the general problem to be “reduced” to sampling from an adaptively constructed sequence of $K$ strongly log-concave (SLC) sub-problems. |
Martin J. Wainwright; |
| 70 | Fast, Parallel, Query-Efficient Binary Classification Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the fundamental classification problem of computing a separating hyperplane for a binary-labeled dataset of size $n$ with normalized $d$-dimensional features. Letting $\Phi \in \mathbb{R}^{n \times d}$ denote the feature matrix and $\gamma$ the margin of the maximum-margin separating hyperplane, we present a randomized algorithm that solves this problem in $\tilde{O}(\gamma^{-2/3}\, \operatorname{nnz}(\Phi) + \gamma^{-2(\omega+1)/3})$-sequential running time (work), $\tilde{O}(\gamma^{-2/3})$-parallel (computational) depth, and accesses $\Phi$ only through $\tilde{O}(\gamma^{-2/3})$-matrix-vector queries (matvecs). |
Ishani Karmarkar; Liam O’Carroll; Aaron Sidford; |
| 71 | High-Dimensional Gaussian Mean Estimation Under Realizable Contamination Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we establish an information–computation gap in the Statistical Query model and, as a consequence, for low-degree polynomial and polynomial-threshold-function algorithms. |
Ilias Diakonikolas; Daniel M. Kane; Thanasis Pittas; |
| 72 | Linear Regression Under Missing or Corrupted Coordinates Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Unlike the clean setting, where the estimation error vanishes as the number of samples grows, the optimal error in these models remains bounded away from zero and depends on the problem parameters. Our main contribution is a characterization of this error, up to constant factors, over essentially the entire parameter range. |
Ilias Diakonikolas; Jelena Diakonikolas; Daniel M. Kane; Jasper C. H. Lee; Thanasis Pittas; |
| 73 | A Quasi-Polynomial Time Mean Estimator Under Mean-Shift Contamination with Unknown Covariance Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In the special case where $\Sigma$ is known to be the identity, prior work gave an algorithm with a near-optimal sample complexity of $\mathrm{poly}(d,2^{\epsilon^{-2}})$ and sample-polynomial time. In this work, we provide a quasi-polynomial time algorithm with sample complexity $2^{\mathrm{poly}(\log d/\epsilon)}$ in the more general unknown covariance case, markedly improving upon the only previously known estimator for this setting that incurs exponential runtime. |
Ilias Diakonikolas; Jingyi Gao; Giannis Iakovidis; Daniel M. Kane; Sihan Liu; Thanasis Pittas; |
| 74 | Second-Order Bounds for $[0,1]$-Valued Regression Via Betting Loss Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We consider the $[0,1]$-valued regression problem in the stochastic setting. |
Yinan Li; Sungjoon Yoon; Ethan Huang; Kwang-Sung Jun; |
| 75 | Continuous Time Policy Evaluation Is Easier with Noisy Dynamics Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we study continuous-time stochastic control problems governed by controlled stochastic differential equations with unknown dynamics. |
Samuel Robertson; Thomas Newton; Csaba Szepesvári; |
| 76 | Avoiding Exp($k^*$) Scaling for Thompson Sampling in Combinatorial Semi-Bandits: From Multiple Seeds to A Single Seed Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Although recent advances have achieved polynomial regret for \emph{linear} rewards, designing an efficient TS algorithm for general, non-linear CMABs remains an open challenge. In this paper, we resolve this open question by proposing \emph{Combinatorial Thompson Sampling with a Single Seed} (\texttt{CTS$^3$}). |
Tianyuan Jin; Heyang Zhao; Vincent Y. F. Tan; Quanquan Gu; |
| 77 | Testing Noise Assumptions of Learning Algorithms Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We pose the following question in computational learning theory: \textit{can we efficiently test whether a training set satisfies the assumptions of a given noise model?} |
Surbhi Goel; Adam Klivans; Konstantinos Stavropoulos; Arsen Vasilyan; |
| 78 | A Unified Lower Bound on The Noisy Query Complexity of Boolean Functions Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the query complexity of Boolean functions $f: {0, 1}^n \rightarrow {0, 1}$ in the noisy query model introduced by Feige, Raghavan, Peleg and Upfal [SICOMP 1994]. |
Yuzhou Gu; Xin Li; Yinzhan Xu; |
| 79 | Open Problem: Is Interaction Necessary for Order-Optimal 1-bit Mean Estimation? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: We ask whether interaction is necessary for order-optimal 1-bit mean estimation over nonparametric finite-moment classes. Adaptive threshold-query protocols achieve the … |
Ivan Lau; Jonathan Scarlett; |
| 80 | A Single Stepsize Suffices for Unprojected Linear TD(0): Simultaneous Robust and Fast Rates Via Polyak–Ruppert Averaging Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study linear TD(0) under Markovian sampling, where data are generated along a single trajectory. |
Wei-Cheng Lee; Francesco Orabona; |
| 81 | Density Estimation for Hellinger Via Minimum-distance Estimators: Mixtures of Gaussians, Log-concave, and More Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the task of density estimation, where we hope to accurately estimate a probability density from $n$ samples. |
Spencer Compton; Jerry Li; |
| 82 | Rate-optimal Community Detection Near The KS Threshold Via Node-robust Algorithms Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study community detection in the \emph{symmetric $k$-stochastic block model}, where $n$ nodes are evenly partitioned into $k$ clusters with intra- and inter-cluster connection probabilities $p$ and $q$, respectively. |
Jingqiu Ding; Yiding Hua; Kasper Lindberg; David Steurer; Aleksandr Storozhenko; |
| 83 | Low-Degree Method Fails to Predict Robust Subspace Recovery Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This success has led to the low-degree conjecture, which posits that this method captures the power and limitations of efficient algorithms for a wide class of high-dimensional statistical problems. We identify a natural and basic hypothesis testing problem in $\mathbb{R}^n$ which is polynomial time solvable, but for which the low-degree polynomial method fails to predict its computational tractability even up to degree $k=n^{\Omega(1)}$. |
He Jia; Aravindan Vijayaraghavan; |
| 84 | Fast Algorithms for Learning A Gaussian Under Halfspace Truncation with Optimal Sample Complexity Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the fundamental problem of learning a high-dimensional Gaussian truncated to an unknown halfspace. |
Haitong Liu; Deepak Narayanan Sridharan; David Steurer; Manuel Wiedmer; |
| 85 | On The Gradient Complexity of Private Optimization with Private Oracles Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the running time, in terms of first order oracle queries, of differentially private empirical/population risk minimization of Lipschitz convex losses. |
Michael Menart; Aleksandar Nikolov; |
| 86 | Optimal Variance-Dependent Regret Bounds for Infinite-Horizon MDPs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Online reinforcement learning in infinite-horizon Markov decision processes (MDPs) remains less theoretically and algorithmically developed than its episodic counterpart, with many algorithms suffering from high “burn-in” costs and failing to adapt to benign instance-specific complexity. In this work, we address these shortcomings for two infinite-horizon objectives: the classical average-reward regret and the $\gamma$-regret. |
Guy Zamir; Matthew Zurek; Yudong Chen; |
| 87 | Open Problem: How Much Overparametrization Is Needed for ALS in Tensor Decomposition? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We pose the open problem of proving convergence to the global optimum for $k=o(r^2)$, or proving that a lower bound on the overparametrized rank of $k=\Omega(r^{1+c})$ for some absolute constant $c>0$ is necessary. |
Dionysis Arvanitakis; Vaidehi Srinivas; Aravindan Vijayaraghavan; |
| 88 | Phase Transition for Stochastic Block Model with More Than $\sqrtn$ Communities Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This breakthrough led them to postulate a new threshold for the many-communities regime $K\geq \sqrt{n}$. In this work, we provide evidence supporting their conjecture: 1- We prove that, for \emph{any graph density}, LDP fail to recover communities below the threshold postulated by Chin et al. (2025) ; 2- We prove that community recovery is possible in polynomial time above the postulated threshold, not only in the \emph{sparse regime} considered in Chin et al. (2025), but also in \emph{moderately sparse regimes}, by counting occurrences of some specific motifs inspired by the LDP analysis. |
Alexandra Carpentier; Christophe Giraud; Nicolas Verzelen; |
| 89 | Wasserstein Policy Learning for Distributional Outcomes Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we study offline policy learning with distribution-valued outcomes, where each potential outcome is a probability measure on $\mathbb{R}$ and the reward is defined through a utility functional applied to the Wasserstein barycenter of induced outcome distributions. |
Yiyan Huang; Cheuk Hang Leung; Qi Wu; Zhiheng Zhang; |
| 90 | Price of Metric Universality in Vector Quantization Is at Most 0.11 Bit Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Information theory demonstrates that an optimal algorithm for reducing precision of $W$ depends on the (second order) statistics of $X$ and requires a careful alignment of vector quantization codebook with PCA directions of $X$ (a process known as “waterfilling allocation”). |
Alina Harbuzova; Or Ordentlich; Yury Polyanskiy; |
| 91 | Tight Sample Complexity Bounds for Entropic Best Policy Identification Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: The improvement we get is due to two main novel technical innovations. We leverage the smoothness properties of the exponential utility to derive sharper concentration bounds, and we propose a new stopping rule that exploits further this tightness to obtain a sample complexity that matches the lower bound. |
Amer Essakine; Claire Vernade; |
| 92 | Distribution-Free Sequential Prediction with Abstentions Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study a sequential prediction problem in which an adversary is allowed to inject arbitrarily many adversarial instances in a stream of i.i.d. instances, but at each round, the learner may also \emph{abstain} from making a prediction without incurring any penalty if the instance was indeed corrupted. |
Jialin Yu; Moïse Blanchard; |
| 93 | Optimal Neural Network Approximation of Smooth Compositional Functions on Sets with Low Intrinsic Dimension Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study approximation and statistical learning properties of deep ReLU networks under structural assumptions that mitigate the curse of dimensionality. |
Thomas Nagler; Sophie Langer; |
| 94 | Data Augmentation: A Fourier Analysis Perspective Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This raises a fundamental question: \emph{Can partial data augmentation achieve the same statistical benefits as full augmentation in terms of generalization and sample complexity?} We develop a general framework for investigating this question using Fourier analysis and the representation theory of finite groups. |
Behrooz Tahmasebi; Melanie Weber; Stefanie Jegelka; |
| 95 | Online Convex Optimization with Sublinear Noisy Probes Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce a unified probing model that generalizes two recent lines of work: sublinear \emph{best-expert} queries in the experts setting, and pairwise (comparison-based) feedback available every round in OCO. |
Simone Di Gregorio; Anupam Gupta; Stefano Leonardi; Matteo Russo; |
| 96 | Omniprediction with Long-Term Constraints Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce and study the problem of online omniprediction with long-term constraints. |
Yahav Bechavod; Jiuyao Lu; Aaron Roth; |
| 97 | Efficient Swap Multicalibration of Elicitable Properties Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Subsequently, we propose an oracle-efficient algorithm which when given access to an online agnostic learner, achieves $\tilde{\mathcal{O}}(T^{\frac{1}{r+1}})$ $\ell_r$-swap multicalibration error with high probability ($r \ge 2$) for a hypothesis class with bounded sequential Rademacher complexity and an elicitable property $\Gamma$. |
Lunjia Hu; Haipeng Luo; Spandan Senapati; Vatsal Sharan; |
| 98 | Simultaneous Blackwell Approachability and Applications to Multiclass Omniprediction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our main result is an extension of the recent binary omniprediction algorithm of Okoroafor et al. (2025) to the multiclass setting, with sample complexity (in statistical settings) or regret horizon (in online settings) $\approx \varepsilon^{-(k+1)}$, for $\varepsilon$-omniprediction in a $k$-class prediction problem. En route to proving this result, we design a framework of potential broader interest for solving Blackwell approachability problems where multiple sets must simultaneously be approached via coupled actions. |
Lunjia Hu; Kevin Tian; Chutong Yang; |
| 99 | How Fast Can You Find A Good Hypothesis? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the computational complexity of achieving statistically optimal sample complexity and approximation constants. |
Anders Aamand; Maryam Aliakbarpour; Justin Y. Chen; Sandeep Silwal; |
| 100 | Nearly Linear-Time User-Level DP-SCO with Optimal Rates Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Current methods, such as those based on DP stochastic gradient descent (SGD), often struggle with high gradient computation complexity or suboptimal utility due to the need to privatize every intermediate iterate. In this work, we introduce a new nearly linear-time algorithm that resolves this trade-off and achieves the optimal excess rates via an adaptive outlier removal framework. |
Badih Ghazi; Ravi Kumar; Daogao Liu; Pasin Manurangsi; |
| 101 | Fixed-Parameter Tractability of Private Synthetic Data Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the problem of generating synthetic data under differential privacy. |
Badih Ghazi; Cristóbal Guzmán; Pritish Kamath; Alexander Knop; Ravi Kumar; Pasin Manurangsi; |
| 102 | Characterizing Online and Private Learnability Under Distributional Constraints Via Generalized Smoothness Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Understanding minimal assumptions that enable learning and generalization is perhaps the central question of learning theory. |
Moïse Blanchard; Abhishek Shetty; Alexander Rakhlin; |
| 103 | The Geometry of Efficient Nonconvex Sampling Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present an efficient algorithm for uniformly sampling from an arbitrary compact body $\mathcal{X} \subset \mathbb{R}^n$ from a warm start under isoperimetry and a natural volume growth condition. |
Santosh S. Vempala; Andre Wibisono; |
| 104 | Learning with Simulators: No Regret in A Computationally Bounded World Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Unfortunately, most results rely heavily on independence (or some proxy thereof) of the data-generating process, while results for strongly dependent data are far more limited. Towards addressing this gap, we introduce the framework of simulatable processes, where the learner has access to a simulator that approximates the distribution generating the data (which may be an arbitrarily complex and dependent process). |
Sasha Voitovych; Abhishek Shetty; Noah Golowich; Alexander Rakhlin; |
| 105 | Learning Ising Models from Evolutions (Extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we revisit the problem of learning the structure and parameters of an Ising model from dynamics. |
Jason Gaitonde; Ankur Moitra; Elchanan Mossel; |
| 106 | Adaptive Learning Rates with Surrogate Probability for Follow-the-Perturbed-Leader Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In contrast, BOBW guarantees for its computationally efficient alternative, follow-the-perturbed-leader (FTPL), remain relatively limited since its optimization-free nature ironically makes the design of adaptive, probability-dependent learning rates non-trivial. To address this challenge, we propose an adaptive learning rate for FTPL by introducing surrogate probability functions that can be computed only from the available quantities, without requiring the exact probabilities. |
Jongyeong Lee; Junya Honda; Shinji Ito; Chansoo Kim; |
| 107 | A Simple, Optimal and Efficient Algorithm for Online Exp-concave Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This paper proposes a simple variant of ONS, called LightONS, which reduces the total runtime to $O(d^2 T + d^\omega \sqrt{T \log T})$ while preserving the optimal regret. |
Yi-Han Wang; Peng Zhao; Zhi-Hua Zhou; |
| 108 | Recovery Thresholds for Hidden Weighted Sparse Graphs (extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We investigate the information-theoretic recovery thresholds for a graph hidden in a randomly weighted complete graph. |
Zhe Hou; Jingcheng Liu; |
| 109 | On The Implicit Regularization of Langevin Dynamics with Projected Noise Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study Langevin dynamics with noise projected onto the directions orthogonal to an isometric group action. This mathematical model is introduced to shed new light on the effects of symmetry on stochastic gradient descent for over-parametrized models. |
Govind Menon; Austin Stromme; Adrien Vacher; |
| 110 | Online Learning for Uninformed Markov Games: Empirical Nash-Value Regret and Non-Stationarity Adaptation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we fully address both limitations. |
Junyan Liu; Haipeng Luo; Zihan Zhang; Lillian J. Ratliff; |
| 111 | Testing for A Hidden Geometry in Random Graphs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we investigate the fundamental problem of detecting a faint geometric signal hidden within an otherwise random graph. |
Amit Silber; Mor Oren-Loberman; Wasim Huleihel; |
| 112 | Spectral Valleys and Sharp Failures in Greedy Determinant Maximization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Although worst-case theory predicts exponentially poor performance, greedy methods are often observed to perform substantially better in practice. This work explains this discrepancy through a finer spectrum-dependent analysis of the greedy algorithm. |
Rajiv Khanna; |
| 113 | Defensive Generation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the problem of efficiently producing, in an online fashion, generative models of scalar, multiclass, and vector-valued outcomes that cannot be falsified on the basis of the observed data and a pre-specified collection of computational tests. |
Gabriele Farina; Juan Carlos Perdomo; |
| 114 | Accelerated Convex Optimization Via Hamiltonian Dynamics with Deterministic Integration Time Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We develop Hamiltonian dynamics-based algorithms for smooth convex optimization that achieve accelerated rates of convergence. |
Xiuyuan Wang; Vishwak Srinivasan; Qiang Fu; Siddharth Mitra; Andre Wibisono; Ashia Wilson; |
| 115 | Last-Iterate Convergence of Randomized Kaczmarz and SGD with Greedy Step Size Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In the proof, we introduce the family of stochastic contraction processes, whose behavior can be described by the evolution of a certain deterministic eigenvalue equation, which we analyze via a careful discrete-to-continuous reduction. |
Michał Dereziński; Xiaoyu Dong; |
| 116 | A Distribution Testing Approach to Clustering Distributions Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the following distribution clustering problem: Given a hidden partition of $k$ distributions into $2$ groups, such that the distributions within each group are the same, and the distributions associated with the clusters are pairwise $\varepsilon$-far in total variation, the goal is to recover the partition. |
Gunjan Kumar; Yash Pote; Jonathan Scarlett; |
| 117 | Dimension Reduction Via Sum-of-Squares and Improved Clustering Algorithms for Non-Spherical Mixtures Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We develop a new approach for clustering non-spherical (i.e., arbitrary component covariances) Gaussian mixture models via a subroutine based on the sum-of-squares method that finds a low-dimensional separation-preserving projection of the input data. |
Prashanti Anderson; Mitali Bafna; Rares-Darius Buhai; Pravesh K. Kothari; David Steurer; |
| 118 | Tight Long-Term Tail Decay of (Clipped) SGD in Non-Convex Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: The study of tail behaviour of \textbf{\texttt{SGD}}-induced processes has been attracting a lot of interest, due to offering strong guarantees with respect to individual runs of … |
Aleksandar Armacki; Dragana Bajović; Dušan Jakovetić; Soummya Kar; Ali H Sayed; |
| 119 | Cloning Is As Hard As Learning for Stabilizer States Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Even when allowing for approximation errors, cloning an arbitrary unknown pure state requires as many initial copies as needed to fully learn the state. Rather than arbitrary unknown states, modern quantum learning theory often considers structured classes of states and exploits such structure to develop learning algorithms that outperform general-state tomography. |
Nikhil Bansal; Matthias C. Caro; Gaurav Mahajan; |
| 120 | Ambiguous Online Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a new variant of online learning that we call “ambiguous online learning". |
Vanessa Kosoy; |
| 121 | Privately Estimating Black-Box Statistics Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work we present a scheme that trades off between statistical efficiency (i.e., how much data is needed) and oracle efficiency (i.e., the number of evaluations). |
Günter Steinke; Thomas Steinke; |
| 122 | Truly Adapting to Adversarial Constraints in Constrained MABs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Specifically, under full feedback we propose an algorithm attaining $\widetilde{\mathcal{O}}(\sqrt{T}+C)$ regret and $\widetilde{\mathcal{O}}(\sqrt{T}+C)$ positive violation, where $C$ quantifies the amount of non-stationarity in the constraints. |
Francesco Emanuele Stradi; Kalana Kalupahana; Matteo Castiglioni; Alberto Marchesi; Nicola Gatti; |
| 123 | Revisiting The (Sub)Optimality of Best-of-N for Inference-Time Alignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We demonstrate that, under minimal conditions on the quality of the reference model and learned reward model, properly tuned BoN is both computationally and statistically optimal in achieving high win-rate, partially explaining its widespread practical success. Because BoN remains susceptible to reward-hacking in this setting, we propose a simple and practical variant that provably eliminates reward-hacking while maintaining optimal statistical performance. |
Ved Sriraman; Adam Block; |
| 124 | Strongly Polynomial Time Complexity of Policy Iteration for $L_∞$ Robust MDPs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we show that a robust policy iteration algorithm runs in strongly-polynomial time for $(s, a)$-rectangular $L_\infty$ RMDPs with a constant (fixed) discount factor, resolving an important algorithmic question. |
Ali Asadi; Krishnendu Chatterjee; Ehsan Goharshady; Mehrdad Karrabi; Alipasha Montaseri; Carlo Pagano; |
| 125 | Tight Bounds for Logistic Regression with Large Stepsize Gradient Descent in Low Dimension Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: With a budget of $T$ iterations, it was recently shown that an accelerated $1/T^2$ rate is possible by choosing a large stepsize $\eta = \Theta(\gamma^2 T)$ (where $\gamma$ is the dataset’s margin) despite the resulting non-monotonicity of the loss. In this paper, we provide a tighter analysis of gradient descent for this problem when the data is two-dimensional: we show that GD with a sufficiently large learning rate $\eta$ finds a point with loss smaller than $\mathcal{O}(1/(\eta \gamma^2 T))$, as long as $T \geq \Omega(n/\gamma + 1/\gamma^2)$, where $n$ is the dataset size. |
Michael Crawshaw; Mingrui Liu; |
| 126 | Optimal Sample Complexity Lower Bounds on Conditional Independence Testing Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: While Canonne et al. derived partial lower bounds for the remaining regimes as well, the problem of fully resolving the sample complexity in all parameters remained open. In this work, we settle these open questions and prove optimal sample complexity lower bounds for both of these problems, thereby completely settling the sample complexities up to polylogarithmic factors. |
Jan Seyfried; Neelkanth Mishra; Sayantan Sen; Marco Tomamichel; |
| 127 | Sample-Efficient Omniprediction for Proper Losses Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the problem of constructing probabilistic predictions that lead to effective decisions when employed by downstream users to inform actions. |
Isaac Gibbs; Ryan J. Tibshirani; |
| 128 | Taming The Monster Every Context: Complexity Measure and Unified Framework for Offline-Oracle Efficient Contextual Bandits Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose an algorithmic framework, Offline Estimation to Decisions (OE2D), that reduces contextual bandit learning with general reward function approximation to offline regression. |
Hao Qin; Chicheng Zhang; |
| 129 | A Perfectly Truthful Calibration Measure Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We design a simple, perfectly and strictly truthful, sound, and complete calibration measure in the batch setting: Averaged Two-Bin Calibration Error (ATB). |
Jason Hartline; Lunjia Hu; Yifan Wu; |
| 130 | Online Realizable Regression and Applications for ReLU Networks Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: Realizable online regression can behave very differently from online classification. Even without any margin or stochastic assumptions, realizability may enforce horizon-free … |
Ilan Doron-Arad; Idan Mehalel; Elchanan Mossel; |
| 131 | Conference on Learning Theory 2026: Preface Related Papers Related Patents Related Grants Related Venues Related Experts View Save |
Steve Hanneke; Tor Lattimore; |
| 132 | Quiet Planting for $k$-SAT, Multiple Solutions of Arbitrary Geometry Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: This work initiates the study of quiet planting with an arbitrary number of solutions, proposing the first method to construct quiet planting distributions for $k$-SAT formulas that accommodate more than one solution. |
Ali Ahmadi; Kiarash Banihashem; Iman Gholami; Mohammad Taghi Hajiaghayi; Jan Olkowski; |
| 133 | Margin in Abstract Spaces Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We begin with a simple margin-based problem in arbitrary metric spaces: concepts are defined by a center point and classify points according to whether their distance lies below $r$ or above $R$. We show that whenever $R>3r$, this class is learnable in \emph{any} metric space. |
Yair Ashlagi; Roi Livni; Shay Moran; Tom Waknine; |
| 134 | Computing Lewis Weights to High Precision Using Local Relative Smoothness Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We provide algorithms that compute $\epsilon$-estimates of the $\ell_p$-Lewis weights of a matrix $A \in \mathbb{R}^{m \times n}$ for $p \geq 4$ using $O(p^2 \log(m/\epsilon))$ rounds of leverage score computation, where $\ell_p$-Lewis weights and leverage scores are both standard measures of row importance. |
Sander Gribling; Aaron Sidford; Chenyi Zhang; |
| 135 | Learning from Equivalence Queries, Revisited Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Within this framework, we study learning from equivalence queries under both full-information and bandit feedback. |
Mark Braverman; Roi Livni; Yishay Mansour; Shay Moran; Kobbi Nissim; |
| 136 | Stochastic Safe Action Model Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we propose an algorithm to learn stochastic planning models where the distribution over the sets of effects for each action has a small support, but the sets may set values to an arbitrary number of attributes. |
Zihao Deng; Brendan Juba; |
| 137 | Randomization for Faster Exact Optimization of Discounted Markov Decision Processes Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We provide faster running times for exactly solving discounted Markov Decision Processes (DMDPs) in strongly polynomial time. |
Andrei Graur; Aaron Sidford; Ta-Wei Tu; |
| 138 | Equivalence of Coarse and Fine-Grained Models for Learning with Distribution Shift Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In particular, we give an efficient black-box reduction from PQ learning to TDS learning for any Boolean concept class. |
Shyamal Patel; Adam Klivans; Konstantinos Stavropoulos; Arsen Vasilyan; |
| 139 | On The Stability of Nonlinear Dynamics in GD and SGD: Beyond Quadratic Potentials Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we explicitly study the effect of nonlinear terms. |
Rotem Mulayoff; Sebastian U. Stich; |
| 140 | Invited Open Problem: Does Differential Privacy Make PAC Learning Much Harder? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: What is the optimal sample complexity of differentially private (DP) PAC learning? Recent results establish that a concept class $C$ is learnable under approximate DP if and only … |
Kobbi Nissim; Uri Stemmer; Eliad Tsfadia; |
| 141 | Online Market Making and The Value of Observing The Order Book Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In the stochastic setting with i.i.d. market prices, we propose an elimination-based algorithm that achieves $\widetilde O(\sqrt{T})$ regret with high probability, without requiring any smoothness assumptions on the distribution of trader valuations. |
Davide Maran; Marcello Restelli; |
| 142 | A Complexity Measure for Active Learning in Multi-group Mean Estimation Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We develop a local minimax framework and prove the first general lower bound for this objective, valid for any finite-variance hypothesis class $\mathcal H$. |
Abdellah Aznag; Rachel Cummings; Adam N. Elmachtoub; |
| 143 | Active Learning on Adversarially Corrupted Graphs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our goal is to design an active learning algorithm that efficiently finds the subset of corrupted vertices using a small number of label queries. |
Marco Bressan; Nicolò Cesa-Bianchi; Tommaso d’Orsi; Emmanuel Esposito; Silvio Lattanzi; |
| 144 | Is Multi-Distribution Learning As Easy As PAC Learning: Sharp Rates with Bounded Label Noise Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Towards understanding the statistical complexity of learning from heterogeneous sources, we study the problem of multi-distribution learning. |
Rafael Hanashiro; Abhishek Shetty; Patrick Jaillet; |
| 145 | An Exponential Lower Bound for Spectral Density Estimation on Unweighted Graphs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: It was left open whether this lower bound could be extended to \emph{unweighted} graphs. In this paper, we answer this question in the affirmative by proving an exponential lower bound for unweighted graphs. |
Pan Peng; Yuyang Wang; Joy Qiping Yang; Yichun Yang; |
| 146 | Robust Algorithms for Finding Cliques in Random Intersection Graphs Via Sum-of-Squares Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work we obtain the first efficient algorithms for recovering the community structure of RIGs both from the perspective of exact and approximate recovery. |
Andreas Göbel; Janosch Ruff; Leon Schiller; |
| 147 | Information-Theoretic Thresholds for Bipartite Latent-Space Graphs Under Noisy Observations Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study information-theoretic phase transitions for the detectability of latent geometry in bipartite random geometric graphs (RGGs) with Gaussian $d$-dimensional latent vectors, while only a subset of edges carries latent information, determined by a random mask with i.i.d. $\mathsf{Bern}(q)$ entries. |
Andreas Göbel; Marcus Pappik; Leon Schiller; |
| 148 | Statistical Learning from Attribution Sets Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We address the problem of training conversion prediction models in advertising domains under privacy constraints, where direct links between ad clicks and conversions are unavailable. |
Lorne Applebaum; Robert Busa-Fekete; August Chen; Claudio Gentile; Tomer Koren; Aryan Mokhtari; |
| 149 | Learning Periodic Strategies in Blocking Bandits Is As Hard As Bandits with Switching Costs Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our main technical contribution is the lower bound, which establishes that blocking bandits are at least as hard as bandits with switching costs. |
Nicolò Cesa-Bianchi; Junya Honda; Yuko Kuroki; Atsushi Miyauchi; Lukas Zierahn; |
| 150 | A Tight Lower Bound for Non-stochastic Multi-armed Bandits with Expert Advice Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We determine the minimax optimal expected regret in the classic non-stochastic multi-armed bandit with expert advice problem, by proving a lower bound that matches the upper bound of [Kale ’14]. |
Zachary Chase; Shinji Ito; Idan Mehalel; |
| 151 | Universality of High-dimensional Scaling Limits of Stochastic Gradient Descent (extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We consider statistical tasks in high dimensions whose loss depends on the data only through its projection into a fixed-dimensional subspace spanned by the parameter vectors and certain ground truth vectors. |
Reza Gheissari; Aukosh Jagannath; |
| 152 | On Randomized Algorithms in Online Strategic Classification Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we provide refined upper and lower bounds for online strategic classification in both the realizable and agnostic settings; our bounds depend on the Littlestone dimension $\mathrm{Ldim}(\mathcal{H})$ of the hypothesis class $\mathcal{H}$ and the maximum degree $\Delta$ of the manipulation graph. |
Chase Hutton; Adam Melrod; Han Shao; |
| 153 | Recursively Enumerably Representable Classes and Computable Versions of The Fundamental Theorem of Statistical Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we investigate the relationship between CPAC learning and recursively enumerable representable (RER) classes, hypothesis classes whose members can be algorithmically listed, in the context of the Fundamental Theorem. |
David Kattermann; Lothar Sebastian Krapp; |
| 154 | Wedge Sampling: Efficient Tensor Completion with Nearly-Linear Sample Complexity Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We introduce \emph{Wedge Sampling}, a new non-adaptive sampling scheme for low-rank tensor completion. |
Hengrui Luo; Anna Ma; Ludovic Stephan; Yizhe Zhu; |
| 155 | Polynomial-time Sampling Despite Disorder Chaos Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We show that with high probability over a random graph $\mathbf{G} \sim G(n,1/2)$: (1) the hardcore model (at fugacity $\lambda = 1$) on $\mathbf{G}$ exhibits disorder chaos, and (2) Glauber dynamics run for $O(n)$ time can approximately sample from the hardcore model on $\mathbf{G}$ (in Wasserstein distance). |
Eric Ma; Tselil Schramm; |
| 156 | Private Linear Regression Via A Down-Sensitivity to Privacy Reduction Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: We present a sample- and time-efficient $(\varepsilon,\delta)$-differentially private (DP) algorithm for $d$-dimensional linear regression with a sample complexity of \[ … |
Ittai Rubinstein; Chris Ge; Samuel B. Hopkins; |
| 157 | The Hidden Cost of Approximation in Online Mirror Descent Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work we initiate a systematic study into inexact OMD, and uncover an intricate relation between regularizer smoothness and robustness to approximation errors. |
Ofir Schlisselberg; Uri Sherman; Tomer Koren; Yishay Mansour; |
| 158 | The Monotonicity of The Franz–Parisi Potential Is Equivalent to Low-Degree MMSE Lower Bounds: Extended Abstract Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we show that for estimation problems the power of low-degree polynomials is governed by the monotonicity of the annealed Franz–Parisi potential for a broad family of Gaussian additive models. |
Konstantinos Tsirkas; Leda Wang; Ilias Zadik; |
| 159 | Spectral Recovery of A Planted Triangle-Dense Subgraph Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: For the recovery of the planted subgraph, we propose a simple spectral algorithm and a semidefinite program, both of which use a graph matrix whose entries are local signed triangle counts. |
Sam van der Poel; Cheng Mao; Benjamin McKenna; |
| 160 | Stable Algorithms Lower Bounds for Estimation from MMSE Discontinuities: Extended Abstract Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we show that for all statistical estimation problems, a natural MMSE instability (discontinuity) condition implies the failure of stable algorithms, serving as a version of OGP for estimation tasks. |
Xifan Yu; Ilias Zadik; |
| 161 | On Efficient Robust Regression with Subquadratic Samples Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We revisit the problem of robust linear regression under Gaussian covariates with an unknown covariance matrix of condition number $\kappa$. |
Deeksha Adil; Jarosław Błasiok; Hongjie Chen; Deepak Narayanan Sridharan; |
| 162 | Optimal Prediction-Augmented Algorithms for Testing Independence of Distributions Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In the non-parametric finite-sample regime, this task is notoriously expensive, as the minimax sample complexity scales polynomially with the support size. In this work, we move beyond these worst-case limitations by leveraging the framework of augmented distribution testing. |
Maryam Aliakbarpour; Alireza Azizi; Ria Stevens; |
| 163 | Variational Tail Bounds for Norms of Random Vectors and Matrices Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a variational tail bound for norms of random vectors and matrices under moment assumptions on their one-dimensional marginals. |
Sohail Bahmani; |
| 164 | Algorithmic Thinking Theory Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this context, a reasoning plan for generating and combining a set of solutions can be thought of as an algorithm for reasoning using a probabilistic oracle. We introduce a theoretical framework for analyzing such reasoning algorithms. |
MohammadHossein Bateni; Vincent Cohen-Addad; Yuzhou Gu; Silvio Lattanzi; Simon Meierhans; Christopher Mohri; |
| 165 | Adaptive Weighted Averaging Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the problem of selecting the largest among $n$ unknown values $x_1,…,x_n$ given only a single unbiased estimate $y_i$ for each $x_i$. |
Aditya Bhaskara; Ashok Cutkosky; Ravi Kumar; Manish Purohit; |
| 166 | Actively Learning Halfspaces Without Synthetic Data Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our objective in this work is to design efficient algorithms for learning halfspaces without point synthesis. |
Hadley Black; Kasper Green Larsen; Arya Mazumdar; Barna Saha; Geelon So; |
| 167 | Eigen-Spike Emergence and Quadratic Equivalents for Conjugate Kernels on Nonlinearly Separable Data Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Recent work in random matrix theory (RMT) has developed the notion of deterministic equivalents: typically linear surrogate models that approximate the spectral behavior of large nonlinear random matrices, such as nonlinear feature maps in neural networks (NNs). Such equivalents make theoretical predictions tractable by reducing a complex model to a simpler one with properties that fall under the umbrella of classical RMT tools. |
Collin Cranston; Zhichao Wang; Todd Kemp; W. Michael Mahoney; |
| 168 | The Matrix-vector Complexity of Ax=b Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Matrix–vector algorithms, particularly Krylov subspace methods, are widely viewed as the most effective algorithms for solving large systems of linear equations. This paper establishes lower bounds on the worst-case number of matrix–vector products needed by such an algorithm to approximately solve a general linear system. |
Michał Dereziński; Ethan N Epperly; Raphael A Meyer; |
| 169 | Minimax Optimal Differentially Private Synthetic Data for Smooth Queries Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose a polynomial-time algorithm that achieves a minimax error rate of $O_{k,d}(n^{-\min {1, \frac{k}{d}}})$, up to a $\log(n)$ factor. |
Rundong Ding; Yiyun He; Yizhe Zhu; |
| 170 | Relatively Smart: A New Approach for Instance-Optimal Learning Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We propose \emph{relatively smart learning}, a new framework which demands that a supervised learner compete only with the best “certifiable” semi-supervised guarantee. |
Shaddin Dughmi; Alireza F. Pour; |
| 171 | The Median Is Easier Than It Looks: Approximation with A Constant-Depth, Linear-Width ReLU Network Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We present depth-width tradeoffs under several settings, culminating in a constant-depth, linear-width construction that achieves exponentially small approximation error with respect to the uniform distribution over the unit hypercube. |
Abhigyan Dutta; Itay Safran; Paul Valiant; |
| 172 | Theoretical Compression Bounds for Wide Multilayer Perceptrons Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: The algorithm we consider bears some similarities with Optimal Brain Damage (OBD) and can be viewed as a post-training randomized version of it. |
Houssam El Cheairi; David Gamarnik; Rahul Mazumder; |
| 173 | Leveraging Similarities in Multi-Armed Bandits Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In many online learning and bandit problems, the actions we consider possess inherent similarities–for instance because they share latent traits, tags, or hierarchical structure.We then provide a unified set of algorithms which adapt to a wide range of richer feedback models, from semi-bandit feedback down to multi-point bandit protocols, including the minimal two-point feedback setting. |
Khaled Eldowa; Thibaud Rahier; Augustin Cablant; Panayotis Mertikopoulos; Pierre Gaillard; |
| 174 | Optimal Reconstruction from Linear Queries Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Our main goal is to characterize the optimal \emph{reconstruction error} as a function of the number of queries $T$, the ambient dimension $d$, and the noise parameter $\delta$. |
Yuval Filmus; Shay Moran; Elizaveta Nesterova; |
| 175 | Fast and Large-Scale Unbalanced Optimal Transport Via Its Semi-Dual and Adaptive Gradient Methods Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, scalable algorithms for UOT, specifically those based on Gradient Descent (SGD), remain largely underexplored. In this work, we address this gap by analyzing the semi-dual formulation of Entropic UOT and demonstrating its suitability for adaptive gradient methods. |
Ferdinand Genans; |
| 176 | High Probability Convergence Guarantees of Stochastic Gradient Descent Ascent in Structured Nonconvex Min-Max Games Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: However, its theoretical foundations remain largely limited to in-expectation convergence guarantees, which fail to capture the failure probability of individual training trajectories, particularly in the presence of heavy-tailed noise. In this work, we bridge this gap by establishing the first high-probability convergence guarantees of stochastic gradient descent-ascent (SGDA) in structured nonconvex games, specifically nonconvex-PL (NC-PL) and nonconvex-concave (NC-C) problems. |
Junsoo Ha; |
| 177 | Learning from Biased and Costly Data Sources: Minimax-optimal Data Collection Under A Budget (extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we study multi-source data collection under a fixed budget, focusing on the estimation of population means and group-conditional means. |
Michael O. Harding; Vikas Singh; Kirthevasan Kandasamy; |
| 178 | Uniform Laws of Large Numbers in Product Spaces Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we study uniform convergence phenomena in \emph{cartesian product spaces}, under assumptions on the underlying distribution that are compatible with the product structure. |
Ron Holzman; Shay Moran; Alexander Shlimovich; |
| 179 | On The Importance of Randomization in Discriminative Feature Feedback Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we show that in sharp contrast to Online Learning, in DFF there can be an unbounded ratio between the optimal mistake bound of deterministic algorithms and that of randomized algorithms, even in the realizable setting. |
Valentio Iverson; Tosca Lechner; Sivan Sabato; |
| 180 | Sandwiching Polynomials for Geometric Concepts with Low Intrinsic Dimension Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we give a new method for constructing low-degree sandwiching polynomials that yield greatly improved degree bounds for several fundamental function classes and marginal distributions. |
Adam Klivans; Konstantinos Stavropoulos; Arsen Vasilyan; |
| 181 | Overlap Analysis of The Shortest Path Problem: Local Search, Landscapes, and Franz-Parisi Potential Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: For optimization problems with random objectives or constraints, ideas originating in statistical physics suggest we should study the \emph{overlap} between approximately-optimal solutions. |
Frederic Koehler; Joonhyung Shin; |
| 182 | Clipping The Price of Adaptivity at The Tail Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Specifically, we assume that the objective decomposes into a model and a loss function, enabling us to intervene by modifying the model’s output before it passes to the loss function. Under this assumption, we design a method that clips the learned model output in tail events where it deviates too much from the output of a fixed reference model. |
Itai Kreisler; Yair Carmon; Oliver Hinder; |
| 183 | How Does The ReLU Activation Affect The Implicit Bias of Gradient Descent on High-dimensional Neural Network Regression? Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this paper, we characterize the implicit bias of GD for training a shallow ReLU model with the squared loss on high-dimensional random features. |
Kuo-Wei Lai; Guanghui Wang; Molei Tao; Vidya Muthukumar; |
| 184 | Blackwell Approachability and Gradient Equilibrium Are Equivalent Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we show that GEQ is equivalent to Blackwell approachability in the algorithmic sense. |
Brian W. Lee; Nika Haghtalab; Michael I. Jordan; Ryan J. Tibshirani; |
| 185 | Self-Concordant Perturbations for Linear Bandits Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Within this framework, we introduce self-concordant perturbations, a family of probability distributions that mirror the role of self-concordant barriers previously employed in the FTRL-based SCRiBLe algorithm. |
Lucas Lévy; Jean-Lou Valeau; Arya Akhavan; Patrick Rebeschini; |
| 186 | On The Power of Adaptivity for $\varepsilon$-Best Arm Identification in Linear Bandits Related Papers Related Patents Related Grants Related Venues Related Experts View Save Abstract: We study the minimax sample complexity of $\varepsilon$-best arm identification in linear bandits, a classical pure-exploration problem. Given a compact action set $\mathcal{X}$ … |
Arnab Maiti; Yunbei Xu; Kevin Jamieson; |
| 187 | Phase Transition in Convex Relaxations for Graph Alignment Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the graph alignment problem for correlated Gaussian Orthogonal Ensemble (GOE) matrices, where the goal is to recover a hidden vertex permutation given two correlated symmetric Gaussian matrices $(A,B)$ with correlation $1/\sqrt{1+\sigma^2}$. |
Laurent Massoulié; Sushil Mahavir Varma; Louis Vassaux; Irène Waldspurger; |
| 188 | Boosting with List-Decodable Codes Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Specifically, we present a new boosting algorithm that, for any class $\mathcal{F}$ closed under $O(\log \frac{1}{\gamma})$-\textsc{Xor}, strong learns $\mathcal{F}$ using $O(\log \frac{1}{\epsilon})$ calls to a $\gamma$-advantage weak learner and a single batch of $\Tilde{O}(\log(\frac{1}{\epsilon})/\gamma^2)$ additional samples. |
Addison Prairie; Li-Yang Tan; |
| 189 | Deep Q-Learning on Hölder Spaces Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We study the operator-theoretic core of Q-learning in continuous-time stochastic control with continuous states and actions. |
Qian Qi; |
| 190 | A Depth Hierarchy for Computing The Maximum in ReLU Networks Via Extremal Graph Theory Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We consider the problem of exact computation of the maximum function over $d$ real inputs using ReLU neural networks. |
Itay Safran; |
| 191 | Finite Sample Bounds for Learning with Score Matching Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work, we provide a non-asymptotic sample complexity analysis for learning the structure of exponential families of polynomials with score matching. |
Devin Smedira; Abhijith Jayakumar; Sidhant Misra; Marc Vuffray; Andrey Y. Lokhov; |
| 192 | Trajectory Data Suffices for Statistically Efficient Policy Evaluation in Fixed-Horizon Offline RL with Linear $q^\pi$-Realizability and Concentrability Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In this work we present a statistically efficient learner for policy evaluation under the same assumptions, with the additional requirement that the behavior policy is known. |
Volodymyr Tkachuk; Csaba Szepesvári; Xiaoqi Tan; |
| 193 | On-Average Stability of Multipass Preconditioned SGD and Effective Dimension Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: When the geometry of the population risk curvature and the geometry of the gradient noise do not match, an aggressive choice that improves one aspect can amplify instability along the other, leading to suboptimal statistical behavior. In this paper we employ \emph{on-average algorithmic stability} to connect generalisation of PSGD to the \emph{effective dimension} that depends on these sources of curvature. |
Simon Vary; Tyler Farghly; Ilja Kuzborskij; Patrick Rebeschini; |
| 194 | Minimax Limits of $k$-Fold Cross-Validation Via Majority Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: Leveraging this analysis, we introduce a minimax framework for cross-validation risk estimation and prove that no empirical risk minimization algorithm can achieve an $O(1/n)$ minimax mean-squared error when the number of folds grows with the number of samples $n$; instead, a lower bound of order $\Omega(\sqrt{k}/n)$ is unavoidable. |
Ido Nachum; Ruediger Urbanke; Thomas Weinberger; |
| 195 | Lyapunov-Based Sample Complexity Analysis for Weakly-Coupled MDPs (extended Abstract) Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: A primary technical contribution of our work is a novel Lyapunov-based analysis framework. |
Wu Tianhao; Matthew Zurek; Weina Wang; Qiaomin Xie; |
| 196 | Worst-case Error Bounds for Online Learning of Smooth Functions Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: We investigate worst-case error for the online learning of real functions that have certain smoothness constraints. |
Weian (Andrew) Xie; |
| 197 | Learning Decision-Sufficient Representations for Linear Optimization Related Papers Related Patents Related Grants Related Venues Related Experts View Save Highlight: In a data-driven regime with i.i.d. costs, we propose a cumulative algorithm that aggregates decision-relevant directions across samples, yielding a stable compression scheme of size at most $d^\star$. |
Yuhan Ye; Saurabh Amin; Asuman Özdağlar; |