CAuSE: Decoding Multimodal Classifiers using Faithful Natural Language Explanation

TL;DR

CAuSE is a novel framework that generates faithful natural language explanations for multimodal classifiers, ensuring interpretability and trustworthiness across various datasets and models.

Contribution

It introduces CAuSE, a causal abstraction-based method trained via interchange intervention, to produce faithful explanations for any pretrained multimodal classifier.

Findings

CAuSE outperforms existing methods on a new causal faithfulness metric.

It generalizes across datasets and models.

Qualitative analysis confirms its interpretability advantages.

Abstract

Multimodal classifiers function as opaque black box models. While several techniques exist to interpret their predictions, very few of them are as intuitive and accessible as natural language explanations (NLEs). To build trust, such explanations must faithfully capture the classifier's internal decision making behavior, a property known as faithfulness. In this paper, we propose CAuSE (Causal Abstraction under Simulated Explanations), a novel framework to generate faithful NLEs for any pretrained multimodal classifier. We demonstrate that CAuSE generalizes across datasets and models through extensive empirical evaluations. Theoretically, we show that CAuSE, trained via interchange intervention, forms a causal abstraction of the underlying classifier. We further validate this through a redesigned metric for measuring causal faithfulness in multimodal settings. CAuSE surpasses other methods on this metric, with qualitative analysis reinforcing its advantages. We perform detailed error analysis to pinpoint the failure cases of CAuSE. For replicability, we make the codes available at https://github.com/newcodevelop/CAuSE