重新审视寻找有影响力的例子的方法
Revisiting Methods for Finding Influential Examples
摘要:
最近提出了几种基于实例的可解释性方法,用于为测试时决策寻找有影响力的训练样本,包括影响函数、TraceIn、代表点选择、Grad-Dot和Grad-Cos。通常,这些方法使用LOO影响(库克距离)作为金标准或使用各种化学方法进行评估。在本文中,我们证明了上述所有方法都是不稳定的,即,对初始化、训练数据的排序和批量大小极其敏感。我们认为,这是一个自然的结果,如何在文献中,例子的影响被假定为独立的模型状态和其他例子-并认为它不是。我们表明,LOO的影响和恶意,因此,穷人的指标来衡量基于实例的解释的质量,而是建议评估这些解释的能力,以检测中毒攻击。此外,我们提供了一个简单而有效的基线来改进上述所有方法,并展示了它如何导致下游任务的显著改进。
英文摘要:
Several instance-based explainability methods for finding influential training examples for test-time decisions have been proposed recently, including Influence Functions, TraceIn, Representer Point Selection, Grad-Dot, and Grad-Cos. Typically these methods are evaluated using LOO influence (Cook's distance) as a gold standard, or using various heuristics. In this paper, we show that all of the above methods are unstable, i.e., extremely sensitive to initialization, ordering of the training data, and batch size. We suggest that this is a natural consequence of how in the literature, the influence of examples is assumed to be independent of model state and other examples -- and argue it is not. We show that LOO influence and heuristics are, as a result, poor metrics to measure the quality of instance-based explanations, and instead propose to evaluate such explanations by their ability to detect poisoning attacks. Further, we provide a simple, yet effective baseline to improve all of the above methods and show how it leads to very significant improvements on downstream tasks.
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。