通过CASAL减少幻觉：对比激活引导用于近似学习
Hallucination reduction with CASAL: Contrastive Activation Steering For Amortized Learning
AI总结：
CASAL通过对比激活引导减少LLM幻觉，提升模型可解释性和实用性，适用于多种模型架构。
摘要：
大型语言模型（LLMs）表现出卓越的能力，但常常出现幻觉，即自信地提供错误答案而不是承认无知。先前的研究表明，模型编码了自身知识的线性表示，且激活引导可以减少幻觉。然而，这些方法需要在推理过程中实时监控和干预。我们介绍了对比激活引导用于近似学习（CASAL），一种高效的算法，将可解释性与近似优化结合起来。CASAL直接将激活引导的好处嵌入模型的权重中。一旦训练完成，LLMs在知道答案时回答问题，而在不知道答案时避免回答。CASAL轻量级的设计只需训练单个transformer层的一个子模块，却在多个短文本问答基准上将幻觉减少了30%-40%。CASAL的计算效率比强LoRA基线如SFT和DPO高30倍，数据效率高20倍，提高了在数据稀缺领域的实用性。重要的是，CASAL也有效地泛化到非分布（OOD）领域。我们展示了CASAL在文本仅和视觉-语言模型中缓解幻觉的灵活性。据我们所知，CASAL是第一个已证明对密集和专家混合（MoE）模型都有效的基于引导的训练方法。CASAL代表了将可解释性启发方法应用于实际部署在生产系统中的重要一步。
英文摘要：
Large Language Models (LLMs) exhibit impressive capabilities but often hallucinate, confidently providing incorrect answers instead of admitting ignorance. Prior work has shown that models encode linear representations of their own knowledge and that activation steering can reduce hallucinations. These approaches, however, require real-time monitoring and intervention during inference. We introduce Contrastive Activation Steering for Amortized Learning (CASAL), an efficient algorithm that connects interpretability with amortized optimization. CASAL directly bakes the benefits of activation steering into model's weights. Once trained, LLMs answer questions they know while abstaining from answering those they do not. CASAL's light-weight design requires training only a submodule of a single transformer layer and yet reduces hallucination by 30%-40% across multiple short-form QA benchmarks. CASAL is 30x more compute-efficient and 20x more data-efficient than strong LoRA-based baselines such as SFT and DPO, boosting its practical applicability in data scarce domains. Importantly, CASAL also generalizes effectively to out-of-distribution (OOD) domains. We showcase CASAL's flexibility in mitigating hallucinations in both text-only and vision-language models. To our knowledge, CASAL is the first steering-based training method that has been shown to be effective for both dense and Mixture-of-Experts (MoE) models. CASAL represents a promising step forward for applying interpretability-inspired method for practical deployment in production systems.
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。
