Annealed Softmax Greedy in Many-Armed Bayesian Bandits 事件

Annealed Softmax Greedy in Many-Armed Bayesian Bandits arXiv:2605.31034v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) and group-based policy optimization methods such as GRPO update a stochastic policy by sampling multiple completions per prompt and increasing the policy's probability on those with higher reward, regularized by a KL penalty toward a reference policy. These updates do not include explicit mechanisms that track epistemic uncertainty. This