{"id":"arxiv/2610.35656","url":"https://arxiv.org/abs/2610.35656","tags":["stat.ML","cs.AI","cs.LG","math.ST"],"score":0,"title":"Provably Efficient Q-learning for Robust Reinforcement Learning","domain":"stat.ML","source":"arxiv","authors":["Yuchen Jiao","Gen Li"],"reviews":0,"subject":"stat.ML","summary":"Standard reinforcement learning (RL) algorithms based on Markov Decision Processes (MDPs) are vulnerable to sim-to-real gaps. To address this issue, distributionally robust RL algorithms based on robust MDPs (RMDPs) have been proposed, for ensuring that agents can perform reliably to unpredictable conditions in real-world environments. Despite of the rich literature, \\revision{an efficient model-free method is still lacking} for robust RL, particularly for Markovian data, which is more relevant in practical applications. In this work, we take a step towards settling this problem by proposing…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-10-26","reactions":{}}