梯度引导的注意力图编辑：迈向高效的上下文幻觉缓解
Gradient-guided Attention Map Editing: Towards Efficient Contextual Hallucination Mitigation
AI总结：
本文提出梯度引导的注意力图编辑方法 GAME，通过分类器识别易诱发幻觉的注意力图并进行梯度引导的权重重分配干预，在摘要与开卷问答任务中有效减少上下文幻觉，并在 XSum 上实现幻觉降低 10% 且计算效率提升 7 倍。
摘要：
在摘要生成和开卷问答（QA）等任务中，大语言模型（LLMs）经常遭遇“上下文幻觉”，即尽管能够获取准确的源信息，却仍生成不相关或错误的响应。这通常是因为这些模型倾向于优先考虑自身生成的内容而非输入上下文，从而导致它们忽略相关细节。为应对这一挑战，我们提出了一种名为“引导注意力图编辑”（GAME）的新方法，该方法动态调整注意力图以提高上下文相关性。在推理过程中，GAME 使用一个训练好的分类器来识别容易诱发幻觉的注意力图，并执行有针对性的干预。这些干预由梯度信息引导的“编辑方向”指导，策略性地在不同注意力头之间重新分配注意力权重，从而有效减少幻觉。在具有挑战性的摘要生成和开卷问答任务上的综合评估表明，GAME 在多种开源模型上持续减少幻觉。具体而言，GAME 在 XSum 摘要任务中将幻觉减少了 10%，同时与最先进的基线方法相比，计算效率实现了 7 倍的加速。
英文摘要：
In tasks like summarization and open-book question answering (QA), Large Language Models (LLMs) often encounter "contextual hallucination", where they produce irrelevant or incorrect responses despite having access to accurate source information. This typically occurs because these models tend to prioritize self-generated content over the input context, causing them to disregard pertinent details. To address this challenge, we introduce a novel method called "Guided Attention Map Editing" (GAME), which dynamically adjusts attention maps to improve contextual relevance. During inference, GAME employs a trained classifier to identify attention maps prone to inducing hallucinations and executes targeted interventions. These interventions, guided by gradient-informed "edit directions'', strategically redistribute attention weights across various heads to effectively reduce hallucination. Comprehensive evaluations on challenging summarization and open-book QA tasks show that GAME consistently reduces hallucinations across a variety of open-source models. Specifically, GAME reduces hallucinations by 10% in the XSum summarization task while achieving a 7X speed-up in computational efficiency compared to the state-of-the-art baselines.
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。
