Spend Your Rollouts Where It Counts: Rollout Allocation for Group-Based RL Post-Training 事件

Spend Your Rollouts Where It Counts: Rollout Allocation for Group-Based RL Post-Training arXiv:2605.26606v1 Announce Type: cross Abstract: Reinforcement learning (RL) is the dominant paradigm for post-training large language models. However, in the online, on-policy setting, rollout generation dominates the computational cost of training. Group-based policy optimization methods compute advantages from multiple rollouts per prompt, yet they indiscriminately allocate budget to prompts with collap