Crafting Large Language Models for Enhanced Interpretability (2407.04307v1)

Abstract: We introduce the Concept Bottleneck LLM (CB-LLM), a pioneering approach to creating inherently interpretable LLMs. Unlike traditional black-box LLMs that rely on post-hoc interpretation methods with limited neuron function insights, CB-LLM sets a new standard with its built-in interpretability, scalability, and ability to provide clear, accurate explanations. This innovation not only advances transparency in LLMs but also enhances their effectiveness. Our unique Automatic Concept Correction (ACC) strategy successfully narrows the performance gap with conventional black-box LLMs, positioning CB-LLM as a model that combines the high accuracy of traditional LLMs with the added benefit of clear interpretability -- a feature markedly absent in existing LLMs.