Don't be fooled: label leakage in explanation methods and the importance of their quantitative evaluation

Abstract

Feature attribution methods identify which features of an input most influence a model's output. Most widely-used feature attribution methods (such as SHAP, LIME, and Grad-CAM) are "class-dependent" methods in that they generate a feature attribution vector as a function of class. In this work, we demonstrate that class-dependent methods can "leak" information about the selected class, making that class appear more likely than it is. Thus, an end user runs the risk of drawing false conclusions when interpreting an explanation generated by a class-dependent method. In contrast, we introduce "distribution-aware" methods, which favor explanations that keep the label's distribution close to its distribution given all features of the input. We introduce SHAP-KL and FastSHAP-KL, two baseline distribution-aware methods that compute Shapley values. Finally, we perform a comprehensive evaluation of seven class-dependent and three distribution-aware methods on three clinical datasets of different high-dimensional data types: images, biosignals, and text.

arXiv commentAISTATS 2023

References in corpus (10)

• HuggingFace's Transformers: State-of-the-art Natural Language Processing
• Axiomatic Attribution for Deep Networks
• Learning Important Features Through Propagating Activation Differences
• Publicly Available Clinical BERT Embeddings
• ClinicalBERT: Modeling Clinical Notes and Predicting Hospital Readmission
• A Benchmark for Interpretability Methods in Deep Neural Networks
• Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methods
• Explaining by Removing: A Unified Framework for Model Explanation
• A Consistent and Efficient Evaluation Strategy for Attribution Methods
• New-Onset Diabetes Assessment Using Artificial Intelligence-Enhanced Electrocardiography