LatentExplainer: Explaining Latent Representations in Deep Generative Models with Multimodal Large Language Models

TL;DR

LatentExplainer is a novel framework that uses multimodal large language models to generate human-understandable explanations of latent variables in deep generative models, improving interpretability and understanding of these complex models.

Contribution

This paper introduces LatentExplainer, the first method to leverage multimodal large language models for explaining latent variables in deep generative models, addressing interpretability challenges.

Findings

Outperforms existing methods in explanation quality

Effectively interprets latent variables across datasets

Enhances model transparency and trustworthiness

Abstract

Deep generative models like VAEs and diffusion models have advanced various generation tasks by leveraging latent variables to learn data distributions and generate high-quality samples. Despite the field of explainable AI making strides in interpreting machine learning models, understanding latent variables in generative models remains challenging. This paper introduces LatentExplainer, a framework for automatically generating semantically meaningful explanations of latent variables in deep generative models. LatentExplainer tackles three main challenges: inferring the meaning of latent variables, aligning explanations with inductive biases, and handling varying degrees of explainability. Our approach perturbs latent variables, interprets changes in generated data, and uses multimodal large language models (MLLMs) to produce human-understandable explanations. We evaluate our proposed method on several real-world and synthetic datasets, and the results demonstrate superior performance in generating high-quality explanations for latent variables. The results highlight the effectiveness of incorporating inductive biases and uncertainty quantification, significantly enhancing model interpretability.