arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2511.21709cs.CLcs.AIcs.IR
收藏

对大语言模型中选择偏差进行量化与缓解:一种可转移的LoRA微调和高效的多数投票方法

Quantifying and Mitigating Selection Bias in LLMs: A Transferable LoRA Fine-Tuning and Efficient Majority Voting Approach

Carnegie Mellon University Africa(卡内基梅隆大学非洲分校);Carnegie Mellon University(卡内基梅隆大学)

Blessed Guda, Lawrence Francis, Gabrial Zencha Ashungafac, Carlee Joe-Wong, Moise Busogi

AI总结:

本文提出了一种可转移的LoRA微调和高效多数投票方法,用于量化并缓解大语言模型中的选择偏差问题。

摘要:

多项选择题(MCQ)回答是一种广泛用于评估大型语言模型(LLMs)性能的方法。然而,LLMs在MCQ任务中常常表现出选择偏差,其选择受到答案位置或选项符号等因素的影响,而非内容本身。这种偏差削弱了MCQ作为评估框架的可靠性。现有的选择偏差度量指标需要答案标签,并测量预测分布与答案分布之间的差异,但无法完全捕捉模型在不同答案排列下的预测一致性。现有的选择偏差缓解策略有显著局限:多数投票虽然有效,但计算成本过高;基于校准的方法需要验证集,且往往无法跨数据集泛化。为解决这些差距,我们提出了三个关键贡献:(1)一种新的无监督、无标签的排列偏差度量指标(PBM),直接量化模型在答案排列下的预测不一致性,提供更精确的选择偏差度量;(2)一种高效的多数投票方法,称为批量问题-上下文KV缓存(BaQCKV),以显著降低计算成本,同时保持偏差缓解的有效性;(3)基于我们提出的指标和BaQCKV的无监督低秩适应(LoRA-1)微调策略,用于缓解选择偏差,提供一种计算高效且能保持模型泛化能力的替代方案。在多个MCQ基准测试中,我们的方法减少了偏差,提高了准确性的一致性,同时最小化了计算成本。

英文摘要:

Multiple Choice Question (MCQ) answering is a widely used method for evaluating the performance of Large Language Models (LLMs). However, LLMs often exhibit selection bias in MCQ tasks, where their choices are influenced by factors like answer position or option symbols rather than the content. This bias undermines the reliability of MCQ as an evaluation framework. Most existing selection bias metrics require answer labels and measure divergences between prediction and answer distributions, but do not fully capture the consistency of a model's predictions across different orderings of answer choices. Existing selection bias mitigation strategies have notable limitations: majority voting, though effective, is computationally prohibitive; calibration-based methods require validation sets and often fail to generalize across datasets. To address these gaps, we propose three key contributions: (1) a new unsupervised label-free Permutation Bias Metric (PBM) that directly quantifies inconsistencies in model predictions across answer permutations, providing a more precise measure of selection bias, (2) an efficient majority voting approach called Batch Question-Context KV caching (BaQCKV), to significantly reduce computational costs while preserving bias mitigation effectiveness, and (3) an unsupervised Low-Rank Adaptation (LoRA-1) fine-tuning strategy based on our proposed metric and the BaQCKV that mitigates selection bias, providing a computationally efficient alternative that maintains model generalizability. Experiments across multiple MCQ benchmarks demonstrate that our approaches reduce bias, increasing consistency in accuracy while minimizing computational costs.

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。

↑