Bootstrapping Exploration with Group-Level Natural Language Feedback in Reinforcement Learning

HuangMeow, Xiang Cheng, Chenxiao Zhao, guobin.shen, Junjie Yang, Xiaocheng Feng, Yuxuan Gu, Xing Yu, Bing Qin

Bootstrapping Exploration with Group-Level Natural Language Feedback in Reinforcement Learning: 38 upvotes on Hugging Face Daily Papers, #1 of 29 papers on 2026-03-12. Day-by-day upvote history. It lost 173 votes when the Hub removed votes in bulk.

Large language models (LLMs) typically receive diverse natural language (NL) feedback through interaction with the environment. However, current reinforcement learning (RL) algorithms rely solely on scalar rewards, leaving the rich information in NL feedback underutilized and leading to inefficient exploration. In this work, we propose GOLF, an RL framework that explicitly exploits group-level language feedback to guide targeted exploration through actionable refinements. GOLF aggregates two complementary feedback sources: (i) external critiques that pinpoint errors or propose targeted fixes, and (ii) intra-group attempts that supply alternative partial ideas and diverse failure patterns. These group-level feedbacks are aggregated to produce high-quality refinements, which are adaptively injected into training as off-policy scaffolds to provide targeted guidance in sparse-reward regions. Meanwhile, GOLF jointly optimizes generation and refinement within a unified RL loop, creating a virtuous cycle that continuously improves both capabilities. Experiments on both verifiable and non-verifiable benchmarks show that GOLF achieves superior performance and exploration efficiency, achieving 2.2times improvements in sample efficiency compared to RL methods trained solely on scalar rewards. Code is available at https://github.com/LuckyyySTA/GOLF.

Paper page on Hugging Face · arXiv

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.