Gnothi Seauton: Empowering Faithful Self-Interpretability in Black-Box Transformers

TL;DR

This paper introduces AutoGnothi, a novel, efficient method that enhances black-box transformers with self-interpretability and accurate explanations, bridging the gap between interpretability and performance in AI models.

Contribution

AutoGnothi is a parameter-efficient pipeline that enables black-box models to generate Shapley value explanations without altering original parameters, improving interpretability and efficiency.

Findings

AutoGnothi provides accurate explanations for vision and language tasks.

It significantly reduces memory, training, and inference costs.

AutoGnothi outperforms traditional parameter-efficient methods.

Abstract

The debate between self-interpretable models and post-hoc explanations for black-box models is central to Explainable AI (XAI). Self-interpretable models, such as concept-based networks, offer insights by connecting decisions to human-understandable concepts but often struggle with performance and scalability. Conversely, post-hoc methods like Shapley values, while theoretically robust, are computationally expensive and resource-intensive. To bridge the gap between these two lines of research, we propose a novel method that combines their strengths, providing theoretically guaranteed self-interpretability for black-box models without compromising prediction accuracy. Specifically, we introduce a parameter-efficient pipeline, AutoGnothi, which integrates a small side network into the black-box model, allowing it to generate Shapley value explanations without changing the original network parameters. This side-tuning approach significantly reduces memory, training, and inference costs, outperforming traditional parameter-efficient methods, where full fine-tuning serves as the optimal baseline. AutoGnothi enables the black-box model to predict and explain its predictions with minimal overhead. Extensive experiments show that AutoGnothi offers accurate explanations for both vision and language tasks, delivering superior computational efficiency with comparable interpretability.