Concept Based Continuous Prompts for Interpretable Text Classification

TL;DR

This paper introduces a concept-based interpretability framework for continuous prompts in text classification, decomposing prompts into human-readable concepts to enhance understanding and maintain performance.

Contribution

It proposes a novel method to interpret continuous prompts by decomposing them into concepts using a concept pool and submodular optimization, improving interpretability without sacrificing accuracy.

Findings

Achieves similar performance to original P-tuning with fewer concepts

Provides more plausible and human-understandable prompt interpretations

Demonstrates the feasibility of concept decomposition for continuous prompts

Abstract

Continuous prompts have become widely adopted for augmenting performance across a wide range of natural language tasks. However, the underlying mechanism of this enhancement remains obscure. Previous studies rely on individual words for interpreting continuous prompts, which lacks comprehensive semantic understanding. Drawing inspiration from Concept Bottleneck Models, we propose a framework for interpreting continuous prompts by decomposing them into human-readable concepts. Specifically, to ensure the feasibility of the decomposition, we demonstrate that a corresponding concept embedding matrix and a coefficient matrix can always be found to replace the prompt embedding matrix. Then, we employ GPT-4o to generate a concept pool and choose potential candidate concepts that are discriminative and representative using a novel submodular optimization algorithm. Experiments demonstrate that our framework can achieve similar results as the original P-tuning and word-based approaches using only a few concepts while providing more plausible results. Our code is available at https://github.com/qq31415926/CD.