Linearly-Interpretable Concept Embedding Models for Text Analysis

TL;DR

This paper introduces LICEM, a novel interpretable text classification model that surpasses previous interpretable models in accuracy, matches black-box models, and can be trained without concept supervision by leveraging LLM backbones.

Contribution

LICEM offers a linearly interpretable concept embedding approach that improves accuracy, enables causal explanations, and reduces annotation needs compared to existing concept-bottleneck models.

Findings

LICEM achieves higher accuracy than existing interpretable models.

LICEM's explanations are more interveneable and causally consistent.

LICEM can be trained without concept supervision using LLM backbones.

Abstract

Despite their success, Large-Language Models (LLMs) still face criticism due to their lack of interpretability. Traditional post-hoc interpretation methods, based on attention and gradient-based analysis, offer limited insights as they only approximate the model's decision-making processes and have been proved to be unreliable. For this reason, Concept-Bottleneck Models (CBMs) have been lately proposed in the textual field to provide interpretable predictions based on human-understandable concepts. However, CBMs still exhibit several limitations due to their architectural constraints limiting their expressivity, to the absence of task-interpretability when employing non-linear task predictors and for requiring extensive annotations that are impractical for real-world text data. In this paper, we address these challenges by proposing a novel Linearly Interpretable Concept Embedding Model (LICEM) going beyond the current accuracy-interpretability trade-off. LICEMs classification accuracy is better than existing interpretable models and matches black-box ones. We show that the explanations provided by our models are more interveneable and causally consistent with respect to existing solutions. Finally, we show that LICEMs can be trained without requiring any concept supervision, as concepts can be automatically predicted when using an LLM backbone.