Zero-shot Concept Bottleneck Models

TL;DR

Zero-shot concept bottleneck models (Z-CBMs) enable interpretable predictions without training by leveraging a large concept bank and cross-modal retrieval, reducing resource needs.

Contribution

The paper introduces Z-CBMs, a novel approach that predicts concepts and labels in a zero-shot manner using a large concept bank and dynamic retrieval.

Findings

Z-CBMs achieve interpretable concept predictions without additional training.

The model effectively retrieves relevant concepts via cross-modal search.

Sparse linear regression selects essential concepts for label inference.

Abstract

Concept bottleneck models (CBMs) are inherently interpretable and intervenable neural network models, which explain their final label prediction by the intermediate prediction of high-level semantic concepts. However, they require target task training to learn input-to-concept and concept-to-label mappings, incurring target dataset collections and training resources. In this paper, we present zero-shot concept bottleneck models (Z-CBMs), which predict concepts and labels in a fully zero-shot manner without training neural networks. Z-CBMs utilize a large-scale concept bank, which is composed of millions of vocabulary extracted from the web, to describe arbitrary input in various domains. For the input-to-concept mapping, we introduce concept retrieval, which dynamically finds input-related concepts by the cross-modal search on the concept bank. In the concept-to-label inference, we apply concept regression to select essential concepts from the retrieved concepts by sparse linear regression. Through extensive experiments, we confirm that our Z-CBMs provide interpretable and intervenable concepts without any additional training. Code will be available at https://github.com/yshinya6/zcbm.