Preserving Reasoning Diversity in Reinforcement Learning with Verifiable Rewards via Adaptive Replay
arXiv:2610.24236v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) improves the reasoning capabilities of large language models, but often reduces the diversity of their reasoning behaviors. We study this phenomenon from the perspective of finite-sample learning d…
Read the full story at arXiv cs.CL ↗
Timeline · 1 report
- 2026-10-21 04:00 · arXiv cs.CL
Preserving Reasoning Diversity in Reinforcement Learning with Verifiable Rewards via Adaptive Replay