Extreme Region Policy Distillation 事件

Extreme Region Policy Distillation arXiv:2605.25582v1 Announce Type: cross Abstract: Reinforcement learning for large language models faces a fundamental trade-off between sample efficiency and asymptotic performance: strictly on-policy methods discard trajectories after a single update, while off-policy reuse introduces distribution mismatch that existing trust-region techniques mitigate primarily by enforcing conservative optimization, often leaving rich training signals underutilized. To inv