Optimistic Thompson Sampling for No-Regret Learning in Unknown Games
1mo ago
Source
IEEEOptimistic Thompson Sampling for No-Regret Learning in Unknown Gamesieee.orgWe study the problem of learning to play a repeated multi-player game with an unknown reward function and bandit feedback. The central challenge arises from the need to balance exploration and exploitation under bandit feedback while strategically responding to other players. To address this, we propose Thompson Sampling-based algorithms that leverage available information about opponents’ actions and reward structures, resulting in a significant reduction in regret bound. Building on these insights, we introduce the Optimism-then-NoRegret framework, which encompasses various game algorithms as special cases. Simulation evaluations on three distinct types of games show that our proposed algorithms consistently and substantially outperform standard baselines.
You might also wanna read
Recurrent Structural Policy Gradient: A Method for Partially Observable Mean Field Games
This article introduces Recurrent Structural Policy Gradient (RSPG), a hybrid structural method for learning history-dependent policies in P
clarisse-wibault.github.io·5d ago
Using Curriculum Learning and PufferLib to Train Superhuman AI Agents for 2048 and Tetris
The article describes using PufferLib, a reinforcement learning framework, to train gaming agents that achieve superhuman performance in 204
kywch.github.io·6mo ago
Contextual Rollout Bandits: A Neural Scheduling Framework for Efficient Reinforcement Learning with Verifiable Rewards
This paper introduces Contextual Rollout Bandits, a novel framework for Reinforcement Learning with Verifiable Rewards (RLVR) that addresses
arxiv.org·1mo ago
New Framework Formalizes Learning from Language Feedback with Provable Performance Guarantees
This paper formalizes the Learning from Language Feedback (LLF) problem, providing a principled framework for interactive learning using lan
arxiv.org·27d ago
Understanding Reinforcement Learning for Model Training, and future directions with GRAPE
arxiv.org·9mo ago
MMR-GRPO: Diversity-Aware Reward Reweighting Accelerates Mathematical Reasoning Model Training
This paper introduces MMR-GRPO, a method that integrates Maximal Marginal Relevance (MMR) into Group Relative Policy Optimization (GRPO) to
arxiv.org·27d ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.