Angles Don’t Lie: Unlocking Training‑Efficient RL Through the Model’s Own Signals

NeurIPSSpotlight2025

Authors
Qinsi Wang, Jinghan Ke, Hancheng Ye, Yueqian Lin, Yuzhe Fu, Jianyi Zhang, Kurt Keutzer, Chenfeng Xu, Yiran Chen
Venue
NeurIPS 2025
Track
Spotlight

TL;DR

Current Reinforcement Fine-tuning (RFT) paradigms for Large Language Models (LLMs) suffer from sample inefficiency due to the redundant exposure of identical queries under uniform data sampling…

Opening excerpt from the authors’ abstract. source

Read the paper

Topics

large language model language model fine-tuning efficient llm

Related NeurIPS papers

← All NeurIPS 2025 Spotlight papers · Browse the whole archive