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 2408.08656cs.CL
收藏

LLMs 对输出格式存在偏见!系统评估与缓解大语言模型的输出格式偏见

LLMs Are Biased Towards Output Formats! Systematically Evaluating and Mitigating Output Format Bias of LLMs

National University of Singapore(新加坡国立大学);VinAI Research(VinAI研究院);Institute of High Performance Computing (IHPC), A*STAR(新加坡科技研究局高性能计算研究所);Salesforce Research(Salesforce研究院);Nanyang Technological University(南洋理工大学)

Do Xuan Long, Hai Nguyen Ngoc, Tiviatis Sim, Hieu Dao, Shafiq Joty, Kenji Kawaguchi, Nancy F. Chen, Min-Yen Kan

AI总结:

本文首次系统评估了LLMs的输出格式偏见,提出区分格式约束内外的评估指标,并通过提示与微调合成格式数据的方法,将ChatGPT包装格式性能方差从235.33降至0.71。

摘要:

我们首次对大型语言模型(LLMs)在格式约束下的性能偏差进行了系统性评估。我们的方法区分了格式约束下评估指标的两类表现,以可靠且准确地评估性能:一类衡量在遵守格式约束时的性能,另一类则评估无论是否遵守约束的性能。随后,我们定义了一个衡量LLMs格式偏差的指标,并建立了减少该偏差的有效策略。接着,我们展示了涵盖四种常用类别——多项选择题问答、包装、列表和映射——共15种广泛使用格式的经验性格式偏差评估。我们在八个生成任务上的评估揭示了最先进的LLMs存在显著的格式偏差。我们进一步发现,提高LLMs跨格式的格式指令遵循能力可能减少格式偏差。基于我们的评估结果,我们研究了使用合成格式数据的提示和微调技术以缓解格式偏差。我们的方法成功地将ChatGPT在包装格式间的性能方差从235.33降至0.71(%$^2$)。

英文摘要:

We present the first systematic evaluation examining format bias in performance of large language models (LLMs). Our approach distinguishes between two categories of an evaluation metric under format constraints to reliably and accurately assess performance: one measures performance when format constraints are adhered to, while the other evaluates performance regardless of constraint adherence. We then define a metric for measuring the format bias of LLMs and establish effective strategies to reduce it. Subsequently, we present our empirical format bias evaluation spanning four commonly used categories -- multiple-choice question-answer, wrapping, list, and mapping -- covering 15 widely-used formats. Our evaluation on eight generation tasks uncovers significant format bias across state-of-the-art LLMs. We further discover that improving the format-instruction following capabilities of LLMs across formats potentially reduces format bias. Based on our evaluation findings, we study prompting and fine-tuning with synthesized format data techniques to mitigate format bias. Our methods successfully reduce the variance in ChatGPT's performance among wrapping formats from 235.33 to 0.71 (%$^2$).

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。

↑