被遗漏的原因与模糊的效应：反事实给神经网络解释带来挑战
Missed Causes and Ambiguous Effects: Counterfactuals Pose Challenges for Interpreting Neural Networks
AI总结：
本文分析反事实因果方法在神经网络可解释性中的两大缺陷：遗漏独立充分原因、反事实依赖不可传递，并给出未来研究建议。
摘要：
可解释性研究将因果性的反事实理论视为理所当然。大多数因果方法依赖于对输入或特定模型组件的激活进行反事实干预，随后观察模型输出 logits 或行为的变化。尽管这比相关性方法产生了更可信的证据，但反事实仍存在关键问题，会以特定且可预测的方式使我们的研究结果产生偏差。具体而言：(i) 反事实理论无法有效捕捉同一效应的多个独立充分原因，这导致我们完全遗漏某些原因；(ii) 神经网络中的反事实依赖关系通常不具有传递性，这使得从神经网络中提取和解释因果图的方法变得复杂。我们讨论了这些挑战对可解释性研究人员的影响，并为未来工作提出了具体建议。
英文摘要：
Interpretability research takes counterfactual theories of causality for granted. Most causal methods rely on counterfactual interventions to inputs or the activations of particular model components, followed by observations of the change in models' output logits or behaviors. While this yields more faithful evidence than correlational methods, counterfactuals nonetheless have key problems that bias our findings in specific and predictable ways. Specifically, (i) counterfactual theories do not effectively capture multiple independently sufficient causes of the same effect, which leads us to miss certain causes entirely; and (ii) counterfactual dependencies in neural networks are generally not transitive, which complicates methods for extracting and interpreting causal graphs from neural networks. We discuss the implications of these challenges for interpretability researchers and propose concrete suggestions for future work.
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。
