Towards Robust Interpretability with Self-Explaining Neural Networks

2018/06/20 by David Alvarez-Melis, Tommi Jaakkola, Alvarez-Melis, David +1 · 33 citations
Computer Science · #Adversarial Robustness in Machine Learning #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine Learning and Data Classification

paper · pdf · doi:10.48550/arxiv.1806.07538

openalex publication_date 2018/06/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Most recent work on interpretability of complex machine learning models has focused on estimating a posteriori explanations for previously trained models around specific predictions. Self-explaining models where interpretability plays a key role already during learning have received much less attention. We propose three desiderata for explanations in general -- explicitness, faithfulness, and stability -- and show that existing methods do not satisfy them. In response, we design self-explaining models in stages, progressively generalizing linear classifiers to complex yet architecturally explicit models. Faithfulness and stability are enforced via regularization specifically tailored to such models. Experimental results across various benchmark datasets show that our framework offers a promising direction for reconciling model complexity and interpretability.

Citations

• A Unified Approach to Interpreting Model Predictions
• Methods for Interpreting and Understanding Deep Neural Networks
• "Why Should I Trust You?": Explaining the Predictions of Any Classifier
• Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps
• Learning Multiple Layers of Features from Tiny Images

Cited by

• Back to the Baseline: Examining Baseline Effects on Explainability Metrics
• ϕ-Table: A Statistical Explanation for Global SHAP
• When, How Long and How Much? Interpretable Neural Networks for Time Series Regression by Learning to Mask and Aggregate
• A Self-explainable Model of Long Time Series by Extracting Informative Structured Causal Patterns
• SX-GeoTree: Self-eXplaining Geospatial Regression Tree Incorporating the Spatial Similarity of Feature Attributions
• Structurally Separated Uncertainty in Supervised Latent Variable Models
• ProfileXAI: User-Adaptive Explainable AI
• Additive Models Explained: A Computational Complexity Approach
• ShapeX: Shapelet-Driven Post Hoc Explanations for Time Series Classification Models
• HattriQ: Designing Integrated Gradients for Feature Attribution in Quantum Machine Learning
• What Do You See? Evaluation of Explainable Artificial Intelligence (XAI) Interpretability through Neural Backdoors
• There is More to Attention: Statistical Filtering Enhances Explanations in Vision Transformers
• Nonparametric Identification of Latent Concepts
• Transparent Visual Reasoning via Object-Centric Agent Collaboration
• Concept Learners for Few-Shot Learning
• Generative causal explanations of black-box classifiers
• Interpretable Anomaly Detection with DIFFI: Depth-based Isolation Forest Feature Importance
• Teaching the Machine to Explain Itself using Domain Knowledge
• Shedding Light on Depth: Explainability Assessment in Monocular Depth Estimation
• XProtoNet: Diagnosis in Chest Radiography with Global and Local Explanations
• Investigating Feature Attribution for 5G Network Intrusion Detection
• Value bounds and Convergence Analysis for Averages of LRP attributions
• Toward a Unified Framework for Debugging Concept-based Models
• What Did You Think Would Happen? Explaining Agent Behaviour Through Intended Outcomes
• Understanding Instance-based Interpretability of Variational Auto-Encoders
• Informative Post-Hoc Explanations Only Exist for Simple Functions
• X-Node: Self-Explanation is All We Need
• Age-Normalized HRV Features for Non-Invasive Glucose Prediction: A Pilot Sleep-Aware Machine Learning Study
• Fairwashing: the risk of rationalization
• Foundations of Interpretable Models
• SUB: Benchmarking CBM Generalization via Synthetic Attribute Substitutions
• SHAP-Guided Regularization in Machine Learning Models
• Pulling Back the Curtain on Deep Networks

Related

• Analyzing the Interpretability Robustness of Self-Explaining Models
• A Comprehensive Survey on Self-Interpretable Neural Networks
• New Faithfulness-Centric Interpretability Paradigms for Natural Language Processing
• How Faithful are Self-Explainable GNNs?
• Explaining, Evaluating and Enhancing Neural Networks' Learned Representations