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 2501.06248cs.LGcs.AIcs.CLecon.GNq-fin.EC
收藏

效用启发的奖励变换改善语言模型的强化学习训练

Utility-inspired Reward Transformations Improve Reinforcement Learning Training of Language Models

Google DeepMind(谷歌DeepMind);Google Research(谷歌研究院);London School of Economics(伦敦政治经济学院)

Roberto-Rafael Maura-Rivero, Chirag Nagpal, Roma Patel, Francesco Visin

AI总结:

针对LLM强化学习训练中线性奖励聚合的缺陷,提出基于Inada条件的效用启发奖励变换,增强低值敏感性、降低高值敏感性,使模型更 helpful 且更少 harmful。

摘要:

当前使用强化学习反馈训练大型语言模型(LLMs)的方法,在训练过程中常常对多个奖励函数的输出进行平均。这忽略了单个奖励维度的关键方面以及奖励间的依赖关系,可能导致生成结果次优。在本工作中,我们展示了奖励的线性聚合存在一些脆弱性,可能导致生成文本出现不期望的属性。随后,我们提出了一种受经济学效用函数理论(特别是Inada条件)启发的奖励函数变换,该变换增强了对低奖励值的敏感性,同时降低了对高奖励值的敏感性。我们将我们的方法与现有的线性聚合奖励的基线方法进行比较,并展示了Inada启发的奖励反馈优于传统的加权平均。我们对方法间的差异进行了定量和定性分析,发现使用Inada变换训练的模型在更有帮助的同时,危害性更小。

英文摘要:

Current methods that train large language models (LLMs) with reinforcement learning feedback, often resort to averaging outputs of multiple rewards functions during training. This overlooks crucial aspects of individual reward dimensions and inter-reward dependencies that can lead to sub-optimal outcomes in generations. In this work, we show how linear aggregation of rewards exhibits some vulnerabilities that can lead to undesired properties of generated text. We then propose a transformation of reward functions inspired by economic theory of utility functions (specifically Inada conditions), that enhances sensitivity to low reward values while diminishing sensitivity to already high values. We compare our approach to the existing baseline methods that linearly aggregate rewards and show how the Inada-inspired reward feedback is superior to traditional weighted averaging. We quantitatively and qualitatively analyse the difference in the methods, and see that models trained with Inada-transformations score as more helpful while being less harmful.

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。

↑