优化类级概率重加权系数以实现公平的提示准确率
Optimizing Class-Level Probability Reweighting Coefficients for Equitable Prompting Accuracy
AI总结:
针对LLMs分类与QA中类别准确率不均的问题,提出一种与模型无关的事后概率重加权方法,通过非线性整数规划优化重加权系数,有效提升了公平性与整体准确率。
摘要:
即使我们在设计 LLMs 以实现一致性和安全性时,它们也常常暴露出预训练数据统计规律性带来的偏差(从不成比例的共现到反映人类认知偏差的刻板关联)。这导致分类和 QA 中存在持续且不均衡的类别准确率。这种各类别准确率差异无法通过架构/训练演进或数据扩展来内在解决,使得事后校正对于公平性能至关重要。为了缓解 LLM 类别准确率不平衡,我们开发了一种事后概率重加权方法,通过突出类别准确率差异的 COBias 指标,直接优化不可微的性能驱动和公平性对齐指标。这种事后偏差缓解方法基于离散优化,采用非线性整数规划(NIP)目标,以及具有理论收敛保证的高效元启发式求解框架。该方法与模型无关,从输出类别概率中学习重加权系数以调整 LLM 推理输出,而无需更新内部权重。评估证明了其有效性:降低 COBias(相对降低 61%),提高整体准确率(相对提升 18%),并在多种提示配置中实现稳健的任务内泛化。
英文摘要:
Even as we engineer LLMs for alignment and safety, they often uncover biases from pre-training data's statistical regularities (from disproportionate co-occurrences to stereotypical associations mirroring human cognitive biases). This leads to persistent, uneven class accuracy in classification and QA. Such per-class accuracy disparities are not inherently resolved by architectural/training evolutions or data scaling, making post-hoc correction essential for equitable performance. To mitigate LLM class accuracy imbalance, we develop a post-hoc probability reweighting method that directly optimizes for non-differentiable performance-driven and fairness-aligned metrics, through a novel COBias metric that highlights disparities in class accuracies. This post-hoc bias mitigation method is grounded in discrete optimization with nonlinear integer programming (NIP) objectives and an efficient metaheuristic solution framework with theoretical convergence guarantees. Operating model-agnostically, it learns reweighting coefficients from output class probabilities to adjust LLM inference outputs without internal weight updates. Evaluations demonstrate its effectiveness: reducing COBias (61% relative reduction), increasing overall accuracy (18% relative increase), and achieving robust within-task generalization across diverse prompt configurations.
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。