Generating Counterfactual Explanations with Natural Language

TL;DR

This paper introduces a method for generating natural language counterfactual explanations for image classification, helping users understand what attributes could change an AI's decision.

Contribution

It proposes a novel approach to produce counterfactual textual explanations by identifying missing evidence that could alter classification outcomes.

Findings

Effective generation of counterfactual explanations demonstrated

Quantitative and qualitative analysis validates the approach

Applicable to fine-grained image classification tasks

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.