Concept Bottleneck Models, ICML 2020

Code for the paper: Discover-then-Name: Task-Agnostic Concept Bottlenecks via Automated Concept Discovery. ECCV 2024.

Generalizable AI predicts immunotherapy outcomes across cancers and treatments

[NeurIPS 24] A new training and evaluation framework for learning interpretable deep vision models and benchmarking different interpretable concept-bottleneck-models (CBMs)

[ICLR 2025 Spotlight] This is the official repository for our paper: ''Enhancing Pre-trained Representation Classifiability can Boost its Interpretability''.

Concept bottleneck models for multiview data with incomplete concept sets

[CVPR 2025] Concept Bottleneck Autoencoder (CB-AE) -- efficiently transform any pretrained (black-box) image generative model into an interpretable generative concept bottleneck model (CBM) with minimal concept supervision, while preserving image quality

Code for the paper "CBVLM: Training-free Explainable Concept-based Large Vision Language Models for Medical Image Classification".

[MICCAI 2024] AdaCBM: An Adaptive Concept Bottleneck Model for Explainable and Accurate Diagnosis

Papers on CBMs with short descriptions of paper's content

Semi-supervised Concept Bottleneck Models (SSCBM)

Web Companion for Generalizable AI predicts immunotherapy outcomes across cancers and treatments

[TMLR, 2025] This is the repository for: Revisiting Discover-then-Name Concept Bottleneck Models: A Reproducibility Study.

This project poses a new methodology for assessing and improving sequential concept bottleneck models (CBMs). The research undertaken in this project builds upon the model proposed by Grange et al., of which I was one of the co-authors.

Official code for the paper "Selective Concept Bottleneck Models Without Predefined Concepts" (TMLR 2025)

A unified library to train Concept Bottleneck Models (CBMs) with state-of-the-art methods (Label-Free, VLG-CBM, LaBo, LM4CV, CB-LLM) and seamless concept integration.