通过负注意力分数对齐纠正大型语言模型中的负偏差
Correcting Negative Bias in Large Language Models through Negative Attention Score Alignment
AI总结：
针对语言模型在复杂推理二元决策中的负偏差问题，提出负注意力分数（NAS）量化该偏差，并设计参数高效的NASA微调方法显著减小精确率与召回率差距且保持泛化能力。
摘要：
二元决策任务（如是非题或答案验证）反映了重要的现实场景，例如用户在特定问题上寻求对其决策正确性的确认。在本研究中，我们观察到语言模型在复杂推理任务的二元决策中表现出负偏差。基于我们的观察和基于注意力的模型动态原理，我们提出了一种负注意力分数（NAS），以系统且定量地表述负偏差。基于NAS，我们识别出那些关注指令中提供的负标记作为二元决策答案候选的注意力头，而不顾提示中的问题，并验证了它们与负偏差的关联。此外，我们提出了负注意力分数对齐（NASA）方法，这是一种参数高效的微调技术，用于处理提取出的负偏差注意力头。来自不同领域推理任务和大型模型搜索空间的实验结果表明，NASA显著减少了由负偏差引起的精确率和召回率之间的差距，同时保持了它们的泛化能力。
英文摘要：
A binary decision task, like yes-no questions or answer verification, reflects a significant real-world scenario such as where users look for confirmation about the correctness of their decisions on specific issues. In this work, we observe that language models exhibit a negative bias in the binary decisions of complex reasoning tasks. Based on our observations and the rationale about attention-based model dynamics, we propose a negative attention score (NAS) to systematically and quantitatively formulate negative bias. Based on NAS, we identify attention heads that attend to negative tokens provided in the instructions as answer candidate of binary decisions, regardless of the question in the prompt, and validate their association with the negative bias. Additionally, we propose the negative attention score alignment (NASA) method, which is a parameter-efficient fine-tuning technique to address the extracted negatively biased attention heads. Experimental results from various domains of reasoning tasks and large model search space demonstrate that NASA significantly reduces the gap between precision and recall caused by negative bias while preserving their generalization abilities.
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。
