CoLiDR: Concept Learning using Aggregated Disentangled Representations

TL;DR

CoLiDR introduces a novel approach that combines disentangled representation learning with concept aggregation, enabling interpretable explanations of deep neural networks by linking generative factors to human-understandable concepts across diverse datasets.

Contribution

The paper proposes CoLiDR, a new method that unifies disentangled representations with concept learning, allowing flexible and interpretable explanations for downstream tasks.

Findings

Successfully aggregates generative factors into concepts

Maintains state-of-the-art performance on concept-based tasks

Demonstrates generalizability to various datasets and concept counts

Abstract

Interpretability of Deep Neural Networks using concept-based models offers a promising way to explain model behavior through human-understandable concepts. A parallel line of research focuses on disentangling the data distribution into its underlying generative factors, in turn explaining the data generation process. While both directions have received extensive attention, little work has been done on explaining concepts in terms of generative factors to unify mathematically disentangled representations and human-understandable concepts as an explanation for downstream tasks. In this paper, we propose a novel method CoLiDR - which utilizes a disentangled representation learning setup for learning mutually independent generative factors and subsequently learns to aggregate the said representations into human-understandable concepts using a novel aggregation/decomposition module. Experiments are conducted on datasets with both known and unknown latent generative factors. Our method successfully aggregates disentangled generative factors into concepts while maintaining parity with state-of-the-art concept-based approaches. Quantitative and visual analysis of the learned aggregation procedure demonstrates the advantages of our work compared to commonly used concept-based models over four challenging datasets. Lastly, our work is generalizable to an arbitrary number of concepts and generative factors - making it flexible enough to be suitable for various types of data.