Paper recorded by Signals 4 on 2026-09-08 in cs.AI. Abstract reproduced from arXiv; link to the original below.
Published 2026-09-08 on arXiv · recorded by Signals 4 on 2026-09-09
Category: cs.AI · 人工智能 · first seen 2026-09-09
In reinforcement learning with verifiable rewards (RLVR) trained with group relative policy optimization (GRPO), the KL-free reward-advantage term studied here depends on within-group reward variation. If all rollouts in a group are correct or all are wrong, their group-relative advantages are identically zero; these zero-advantage silent groups provide no reward-advantage gradient, yet uniform sa