Csaba Szepesvári

University of Alberta — Canada

Decision Sciences · Social Sciences

57h-index16.8kcitations475works0.22yr avg
OpenAlex↗ORCID↗Google Scholar↗Lab page↗

Accepting Students?

No reports yet. Know if this professor is taking students?

Research Topics

Advanced Bandit Algorithms Research(230), Reinforcement Learning in Robotics(174), Machine Learning and Algorithms(145), Optimization and Search Problems(59), Auction Theory and Applications(34)

Publications475 total

Sample-Based Planning and Learning with Function Approximation
Statistical Science·2025
Trajectory Data Suffices for Statistically Efficient Policy Evaluation in Finite-Horizon Offline RL with Linear $q^π$-Realizability and Concentrability
arXiv (Cornell University)·2025Open Access
Rectifying Regression in Reinforcement Learning
arXiv (Cornell University)·2025Open Access
Frontier LLMs Still Struggle with Simple Reasoning Tasks
arXiv (Cornell University)·2025Open Access
Beyond Statistical Learning: Exact Learning Is Essential for General Intelligence
arXiv (Cornell University)·2025Open Access
Rethinking the Global Convergence of Softmax Policy Gradient with Linear Function Approximation
arXiv (Cornell University)·2025Open Access
Ordering-based Conditions for Global Convergence of Policy Gradient Methods
arXiv (Cornell University)·2025Open Access
Small steps no more: Global convergence of stochastic gradient bandits for arbitrary learning rates
arXiv (Cornell University)·2025Open Access
Almost Free: Self-concordance in Natural Exponential Families and an Application to Bandits
arXiv (Cornell University)·2024Open Access
Confident Natural Policy Gradient for Local Planning in $q_π$-realizable Constrained MDPs
arXiv (Cornell University)·2024Open Access
To Believe or Not to Believe Your LLM
arXiv (Cornell University)·2024· 3 citedOpen Access
Trajectory Data Suffices for Statistically Efficient Learning in Offline RL with Linear $q^π$-Realizability and Concentrability
arXiv (Cornell University)·2024Open Access
Mitigating LLM Hallucinations via Conformal Abstention
arXiv (Cornell University)·2024· 2 citedOpen Access
Regret Minimization via Saddle Point Optimization
arXiv (Cornell University)·2024Open Access
Switching the Loss Reduces the Cost in Batch (Offline) Reinforcement Learning
arXiv (Cornell University)·2024Open Access
Stochastic Gradient Succeeds for Bandits
arXiv (Cornell University)·2024Open Access
Trajectory Data Suffices for Statistically Efficient Learning in Offline RL with Linear $q^\pi$-Realizability and Concentrability
2024
Confident Natural Policy Gradient for Local Planning in $q_\pi$-realizable Constrained MDPs
2024
Ensemble sampling for linear bandits: small ensembles suffice
2024
Small steps no more: Global convergence of stochastic gradient bandits for arbitrary learning rates
2024
page 1 of 24Next →

Frequent Co-authors

Rémi Munos(9), Dávid Pál(5), Richard S. Sutton(5), András Antos(5), Gergely Neu(5), Levente Kocsis(4), Jean-Yves Audibert(4), Hamid Reza Maei(4), David Silver(4), Gábor Bartók(3), Branislav Kveton(3), András György(3), Yasin Abbasi-Yadkori(2), Tor Lattimore(2), Michael L. Littman(2), Doina Precup(2), Shalabh Bhatnagar(2), Ruitong Huang(2), Sébastien Bubeck(2), Gilles Stoltz(2)