Controllable Concept Bottleneck Models

2026-01-01

arXiv.org

AI Summary

Existing concept bottleneck models typically require full retraining to accommodate dynamic editing demands such as data deletion, concept correction, or incremental addition of new samples, resulting in significant inefficiency. This work proposes the Controllable Concept Bottleneck Model (CCBM), which introduces—for the first time—a closed-form update mechanism enabling three granularities of edits: at the concept label level, concept level, and data level (including additions and deletions). Leveraging influence functions, CCBM derives efficient approximations that allow multi-granular model updates without retraining. Experimental results demonstrate that CCBM maintains interpretability while substantially improving model maintainability and adaptability, dramatically reducing computational overhead and offering strong practical deployment potential.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

Concept Bottleneck Models (CBMs) have garnered much attention for their ability to elucidate the prediction process through a human-understandable concept layer. However, most previous studies focused on static scenarios where the data and concepts are assumed to be fixed and clean. In real-world applications, deployed models require continuous maintenance: we often need to remove erroneous or sensitive data (unlearning), correct mislabeled concepts, or incorporate newly acquired samples (incremental learning) to adapt to evolving environments. Thus, deriving efficient editable CBMs without retraining from scratch remains a significant challenge, particularly in large-scale applications. To address these challenges, we propose Controllable Concept Bottleneck Models (CCBMs). Specifically, CCBMs support three granularities of model editing: concept-label-level, concept-level, and data-level, the latter of which encompasses both data removal and data addition. CCBMs enjoy mathematically rigorous closed-form approximations derived from influence functions that obviate the need for retraining. Experimental results demonstrate the efficiency and adaptability of our CCBMs, affirming their practical value in enabling dynamic and trustworthy CBMs.

Problem

• Concept Bottleneck Models
• model editing
• unlearning
• incremental learning
• influence functions

Innovation

• Concept Bottleneck Models
• Model Editing
• Influence Functions
• Unlearning
• Incremental Learning