CUBIC: Concept Embeddings for Unsupervised Bias Identification using VLMs

TL;DR

CUBIC is a novel method that automatically identifies biases in vision models by analyzing concept influences in the image-text latent space, without needing annotated bias examples or predefined bias concepts.

Contribution

CUBIC introduces an unsupervised approach leveraging vision-language models and linear probes to discover bias-inducing concepts without prior bias annotations or failure examples.

Findings

CUBIC uncovers previously unknown biases in vision models.

It operates without requiring dataset failure samples or bias annotations.

The method effectively identifies influential concepts affecting model predictions.

Abstract

Deep vision models often rely on biases learned from spurious correlations in datasets. To identify these biases, methods that interpret high-level, human-understandable concepts are more effective than those relying primarily on low-level features like heatmaps. A major challenge for these concept-based methods is the lack of image annotations indicating potentially bias-inducing concepts, since creating such annotations requires detailed labeling for each dataset and concept, which is highly labor-intensive. We present CUBIC (Concept embeddings for Unsupervised Bias IdentifiCation), a novel method that automatically discovers interpretable concepts that may bias classifier behavior. Unlike existing approaches, CUBIC does not rely on predefined bias candidates or examples of model failures tied to specific biases, as such information is not always available. Instead, it leverages image-text latent space and linear classifier probes to examine how the latent representation of a superclass label$\unicode{x2014}$shared by all instances in the dataset$\unicode{x2014}$is influenced by the presence of a given concept. By measuring these shifts against the normal vector to the classifier's decision boundary, CUBIC identifies concepts that significantly influence model predictions. Our experiments demonstrate that CUBIC effectively uncovers previously unknown biases using Vision-Language Models (VLMs) without requiring the samples in the dataset where the classifier underperforms or prior knowledge of potential biases.