Bi-ICE: An Inner Interpretable Framework for Image Classification via Bi-directional Interactions between Concept and Input Embeddings

TL;DR

This paper introduces Bi-ICE, a novel framework that enhances interpretability in large-scale image classification by enabling bi-directional interactions between concept and input embeddings, providing insights into model decision processes.

Contribution

The paper proposes the Bi-ICE module supporting multilevel interpretability and demonstrates its effectiveness in revealing concept contributions and localization in image classification.

Findings

Improved transparency in image classification models.

Effective localization of concepts within input images.

Demonstrated interpretability across multiple levels.

Abstract

Inner interpretability is a promising field aiming to uncover the internal mechanisms of AI systems through scalable, automated methods. While significant research has been conducted on large language models, limited attention has been paid to applying inner interpretability to large-scale image tasks, focusing primarily on architectural and functional levels to visualize learned concepts. In this paper, we first present a conceptual framework that supports inner interpretability and multilevel analysis for large-scale image classification tasks. Specifically, we introduce the Bi-directional Interaction between Concept and Input Embeddings (Bi-ICE) module, which facilitates interpretability across the computational, algorithmic, and implementation levels. This module enhances transparency by generating predictions based on human-understandable concepts, quantifying their contributions, and localizing them within the inputs. Finally, we showcase enhanced transparency in image classification, measuring concept contributions, and pinpointing their locations within the inputs. Our approach highlights algorithmic interpretability by demonstrating the process of concept learning and its convergence.