主动层对比解码减少大语言模型生成中的幻觉
Active Layer-Contrastive Decoding Reduces Hallucination in Large Language Model Generation
AI总结：
提出主动层对比解码（ActLCD），通过强化学习策略在生成过程中动态决定何时应用对比层，在五个基准上超越现有方法，有效缓解大语言模型的长上下文幻觉问题。
摘要：
近期解码方法通过改进生成过程中下一个 token 的选择方式，提升了大型语言模型（LLMs）的事实性。这些方法通常在 token 层面运作，利用内部表示来抑制表面模式。然而，LLMs 仍然容易产生幻觉，尤其是在较长上下文中。本文提出主动层对比解码（ActLCD），这是一种新颖的解码策略，能在生成过程中主动决定何时应用对比层。通过将解码转化为序列决策问题，ActLCD 采用由奖励感知分类器引导的强化学习策略，以在 token 层面之外优化事实性。实验表明，ActLCD 在五个基准上超越了最先进的方法，展示了其在多种生成场景中缓解幻觉的有效性。
英文摘要：
Recent decoding methods improve the factuality of large language models (LLMs) by refining how the next token is selected during generation. These methods typically operate at the token level, leveraging internal representations to suppress superficial patterns. Nevertheless, LLMs remain prone to hallucinations, especially over longer contexts. In this paper, we propose Active Layer-Contrastive Decoding (ActLCD), a novel decoding strategy that actively decides when to apply contrasting layers during generation. By casting decoding as a sequential decision-making problem, ActLCD employs a reinforcement learning policy guided by a reward-aware classifier to optimize factuality beyond the token level. Our experiments demonstrate that ActLCD surpasses state-of-the-art methods across five benchmarks, showcasing its effectiveness in mitigating hallucinations in diverse generation scenarios.
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。
