Generating Counterfactual Explanations with Natural Language
arXiv:1806.09809
Abstract
Natural language explanations of deep neural network decisions provide an intuitive way for a AI agent to articulate a reasoning process. Current textual explanations learn to discuss class discriminative features in an image. However, it is also helpful to understand which attributes might change a classification decision if present in an image (e.g., "This is not a Scarlet Tanager because it does not have black wings.") We call such textual explanations counterfactual explanations, and propose an intuitive method to generate counterfactual explanations by inspecting which evidence in an input is missing, but might contribute to a different classification decision if present in the image. To demonstrate our method we consider a fine-grained image classification task in which we take as input an image and a counterfactual class and output text which explains why the image does not belong to a counterfactual class. We then analyze our generated counterfactual explanations both qualitatively and quantitatively using proposed automatic metrics.
presented at 2018 ICML Workshop on Human Interpretability in Machine Learning (WHI 2018), Stockholm, Sweden
Cited by in corpus (12)
- The Hidden Assumptions Behind Counterfactual Explanations and Principal Reasons
- Explaining Classifiers with Causal Concept Effect (CaCE)
- How model accuracy and explanation fidelity influence user trust
- Generating Counterfactual and Contrastive Explanations using SHAP
- The effectiveness of feature attribution methods and its correlation with automatic evaluation scores
- Causal Interpretability for Machine Learning -- Problems, Methods and Evaluation
- Explainable artificial intelligence (XAI), the goodness criteria and the grasp-ability test
- Human-interpretable model explainability on high-dimensional data
- Interpretable and Personalized Apprenticeship Scheduling: Learning Interpretable Scheduling Policies from Heterogeneous User Demonstrations
- SCOUT: Self-aware Discriminant Counterfactual Explanations
- Modularized Textual Grounding for Counterfactual Resilience
- Learning Models for Actionable Recourse