arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.09692cs.LGcs.AIcs.CL
收藏

ActiveUltraFeedback:使用主动学习的高效偏好数据生成

ActiveUltraFeedback: Efficient Preference Data Generation using Active Learning

University of Freiburg(弗赖堡大学)

Davit Melikidze, Marian Schneider, Jessica Lam, Martin Wertich, Ido Hakimi, Barna Pásztor, Andreas Krause

AI总结:

提出ActiveUltraFeedback主动学习流水线,通过不确定性估计和两种新采样方法(DRTS和DeltaUCB)动态选择最具信息量的响应对,以最少六分之一的标注数据实现与静态基线相当或更优的下游性能。

摘要:

基于人类反馈的强化学习(RLHF)已成为对齐大型语言模型(LLMs)的标准方法,但其有效性受到偏好数据获取高成本的瓶颈限制,尤其是在低资源和专家领域。为解决这一问题,我们引入了ACTIVEULTRAFEEDBACK,一个模块化的主动学习流水线,利用不确定性估计动态识别最具信息量的响应进行标注。我们的流水线支持系统评估标准响应选择方法以及两种新方法:DOUBLE REVERSE THOMPSON SAMPLING(DRTS)和DELTAUCB,这两种方法优先选择预测质量差距大的响应对,利用近期研究结果,即此类对为微调提供良好信号。实验表明,ACTIVEULTRAFEEDBACK生成的高质量数据集在下游性能上带来显著提升,尤其以静态基线六分之一的标注数据即可达到相当或更优的结果。我们的流水线可在https://github.com/lasgroup/ActiveUltraFeedback获取,偏好数据集可在https://huggingface.co/ActiveUltraFeedback获取。

英文摘要:

Reinforcement Learning from Human Feedback (RLHF) has become the standard for aligning Large Language Models (LLMs), yet its efficacy is bottlenecked by the high cost of acquiring preference data, especially in low-resource and expert domains. To address this, we introduce ACTIVEULTRAFEEDBACK, a modular active learning pipeline that leverages uncertainty estimates to dynamically identify the most informative responses for annotation. Our pipeline facilitates the systematic evaluation of standard response selection methods alongside DOUBLE REVERSE THOMPSON SAMPLING (DRTS) and DELTAUCB, two novel methods prioritizing response pairs with large predicted quality gaps, leveraging recent results showing that such pairs provide good signals for fine-tuning. Our experiments demonstrate that ACTIVEULTRAFEEDBACK yields high-quality datasets that lead to significant improvements in downstream performance, notably achieving comparable or superior results with as little as one-sixth of the annotated data relative to static baselines. Our pipeline is available at https://github.com/lasgroup/ActiveUltraFeedback and our preference datasets at https://huggingface.co/ActiveUltraFeedback.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑