---
title: "Concept Bottleneck Models for Explainable Decision Making: A Survey of Progress, Taxonomy, and Future Directions"
authors:
  - "Chunjiang Wang"
  - "Fan Li"
  - "Wenbo Hu"
  - "Rui Yan"
  - "Kun Zhang"
  - "Shaohua Kevin Zhou"
doi: "10.24963/ijcai.2026/892"
url: "https://ijcai.org/proceedings/2026/892"
pdf: "https://ijcai.org/proceedings/2026/0892.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "8041-8049"
track: "Survey Track"
keywords:
  - "AI Ethics, Trust, Fairnes: Explainability and interpretability"
---

# Concept Bottleneck Models for Explainable Decision Making: A Survey of Progress, Taxonomy, and Future Directions

Chunjiang Wang, Fan Li, Wenbo Hu, Rui Yan, Kun Zhang, Shaohua Kevin Zhou

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 8041-8049. DOI: [10.24963/ijcai.2026/892](https://doi.org/10.24963/ijcai.2026/892)

## Abstract

Deep neural networks deliver strong performance but remain opaque, limiting their use in high-stakes domains that require transparency and human oversight. Concept Bottleneck Models (CBMs) address this gap by introducing a human-interpretable concept layer that mediates inputs and decisions, enabling semantic explanations and test-time intervention. This survey provides a unified review of CBMs organized along four dimensions: concept acquisition, concept-based decision making, concept intervention, and concept evaluation. We summarize the evolution of concept construction from manual annotation to lexicon-based mining, LLM/VLM-guided generation, and visually grounded discovery via prototypes and diffusion models; review emerging CBM architectures beyond strict bottlenecks; and consolidate evaluation and intervention protocols emphasizing faithfulness, sparsity, and intervenability, with particular relevance to high-stakes domains such as healthcare. We synthesize fragmented literature and outline key challenges and future directions for concept-based interpretable decision making.