A Critical Look at Power Sampling when Test Time Sharpening LLMs
arXiv:2610.13653v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has substantially improved LLM reasoning, yet growing evidence suggests it mainly performs distribution sharpening: concentrating probability mass on reasoning paths the base model already produces…
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-10-12 04:00 · arXiv stat.ML
A Critical Look at Power Sampling when Test Time Sharpening LLMs