重新审视证据深度学习中异常检测的空虚性
Rethinking Vacuity for OOD Detection in Evidential Deep Learning
AI总结:
研究指出证据深度学习中空虚性指标对类别数敏感,通过实验证明类别数差异会导致AUROC和AUPR显著变化,提出需明确ID和OOD定义以避免评估偏差。
摘要:
空虚性或不确定性质量(UM)是评估证据深度学习(EDL)中异常检测的常用指标。它通常通过将类别数(K)除以模型预测的总信念强度(S)来计算,其中S来源于Dirichlet参数之和。因此,UM对K的基数敏感。特别是,当K和S增加时,由于EDL的性质(抑制错误分配的证据),K和S之间不太可能存在线性关系。因此,在比较内分布(ID)和外分布(OOD)结果时,重要的是K_ID和K_OOD相等;这一点在实践中并不总是得到保证。我们提供了实证证明,当ID和OOD之间的类别数差异为1时,AUROC和AUPR的结果可以显著不同,对于标准EDL,AUROC差异可达0.318,AUPR差异达0.613;对于IB-EDL,AUROC差异达0.360,AUPR差异达0.683。更具体地说,我们的发现揭示了一个评估偏差:当ID和OOD之间的K不同时,AUROC/AUPR可以被人为夸大,而模型预测本身没有变化。我们进一步讨论了使用多选问题-答案(MCQA)数据集评估EDL在因果语言模型上的表现,并论证了在此背景下ID和OOD定义的明确性。我们的主要贡献是实证和理论证明,基于空虚性的EDL微调大语言模型的异常检测对未控制的评估类别数差异高度敏感。
英文摘要:
Vacuity, or Uncertainty Mass (UM), is commonly used as a metric to evaluate Out-of-Distribution (OOD) detection in Evidential Deep Learning (EDL). It generally involves dividing the number of classes ($K$) by the total strength of belief ($S$) of the model's predictions, where $S$ is derived from summing the Dirichlet parameters. As such, UM is sensitive to the cardinality of $K$. As a result, when comparing In Distribution (ID) and OOD results, it is important that $K_{\mathrm{ID}}$ and $K_{\mathrm{OOD}}$ are equal; something that is not always ensured in practice. We provide an empirical demonstration of how results for AUROC and AUPR can substantially differ when class cardinality between ID and OOD differs by 1, with AUROC differing by as much as 0.346 and AUPR by 0.634 for standard EDL, and AUROC by 0.427 and AUPR by 0.745 for IB-EDL (both from Implementation B, Llama3-8B, ARC-E). Our findings isolate an evaluation artefact: when $K$ differs between ID and OOD, AUROC/AUPR can be artificially inflated without any change in model predictions. We further discuss the evaluation of EDL over causal language models using Multiple-Choice Question-Answer (MCQA) datasets and argue for clearer definitions of ID and OOD in this context. Our primary contribution is an empirical and theoretical demonstration that vacuity-based OOD detection in EDL-fine-tuned LLMs is highly sensitive to uncontrolled differences in evaluated class cardinality.
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。