Saved in:
Bibliographic Details
Main Authors: Shen, Matthew, Hsu, Aliyah, Agarwal, Abhineet, Yu, Bin
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.06730
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908318348345344
author Shen, Matthew
Hsu, Aliyah
Agarwal, Abhineet
Yu, Bin
author_facet Shen, Matthew
Hsu, Aliyah
Agarwal, Abhineet
Yu, Bin
contents Concept bottleneck models (CBM) aim to improve model interpretability by predicting human level "concepts" in a bottleneck within a deep learning model architecture. However, how the predicted concepts are used in predicting the target still either remains black-box or is simplified to maintain interpretability at the cost of prediction performance. We propose to use Fast Interpretable Greedy Sum-Trees (FIGS) to obtain Binary Distillation (BD). This new method, called FIGS-BD, distills a binary-augmented concept-to-target portion of the CBM into an interpretable tree-based model, while maintaining the competitive prediction performance of the CBM teacher. FIGS-BD can be used in downstream tasks to explain and decompose CBM predictions into interpretable binary-concept-interaction attributions and guide adaptive test-time intervention. Across 4 datasets, we demonstrate that our adaptive test-time intervention identifies key concepts that significantly improve performance for realistic human-in-the-loop settings that only allow for limited concept interventions.
format Preprint
id arxiv_https___arxiv_org_abs_2503_06730
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Adaptive Test-Time Intervention for Concept Bottleneck Models
Shen, Matthew
Hsu, Aliyah
Agarwal, Abhineet
Yu, Bin
Machine Learning
Concept bottleneck models (CBM) aim to improve model interpretability by predicting human level "concepts" in a bottleneck within a deep learning model architecture. However, how the predicted concepts are used in predicting the target still either remains black-box or is simplified to maintain interpretability at the cost of prediction performance. We propose to use Fast Interpretable Greedy Sum-Trees (FIGS) to obtain Binary Distillation (BD). This new method, called FIGS-BD, distills a binary-augmented concept-to-target portion of the CBM into an interpretable tree-based model, while maintaining the competitive prediction performance of the CBM teacher. FIGS-BD can be used in downstream tasks to explain and decompose CBM predictions into interpretable binary-concept-interaction attributions and guide adaptive test-time intervention. Across 4 datasets, we demonstrate that our adaptive test-time intervention identifies key concepts that significantly improve performance for realistic human-in-the-loop settings that only allow for limited concept interventions.
title Adaptive Test-Time Intervention for Concept Bottleneck Models
topic Machine Learning
url https://arxiv.org/abs/2503.06730