Tree of Concepts: Interpretable Continual Learners in Non-Stationary Clinical Domains

2026-04-18

AI Summary

This work addresses the challenge of balancing distribution shifts and interpretability in continual learning under non-stationary clinical settings. The authors propose an interpretable continual learning framework based on concept bottlenecks, introducing for the first time a structured concept interface into this domain. Specifically, they define a stable concept interface using shallow decision trees with fixed semantics, updating only the concept extractor and label head while employing a replay-augmented training strategy. Evaluated on multiple medical tabular data benchmarks, the method significantly outperforms existing baselines, effectively mitigating explanation drift in dynamic environments and achieving a superior trade-off between stability and plasticity.

Technology Category

• Machine Learning: Life-Long and Continual Learning
• Computer Vision: Interpretability, Explainability, and Transparency
• Humans and AI: Explainable AI (XAI) for Human Understanding

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Continual learning aims to update models under distribution shift without forgetting, yet many high-stakes deployments, such as healthcare, also require interpretability. In practice, models that adapt well (e.g., deep networks) are often opaque, while models that are interpretable (e.g., decision trees) are brittle under shift, making it difficult to achieve both properties simultaneously. In response, we propose Tree of Concepts, an interpretable continual learning framework that uses a shallow decision tree to define a fixed, rule-based concept interface and trains a concept bottleneck model to predict these concepts from raw features. Continual updates act on the concept extractor and label head while keeping concept semantics stable over time, yielding explanations that do not drift across sequential updates. On multiple tabular healthcare benchmarks under continual learning protocols, our method achieves a stronger stability-plasticity trade-off than existing baselines, including replay-enhanced variants. Our results suggest that structured concept interfaces can support continual adaptation while preserving a consistent audit interface in non-stationary, high-stakes domains.

Problem

• continual learning
• interpretability
• distribution shift
• clinical domains
• concept drift

Innovation

• interpretable continual learning
• concept bottleneck model
• decision tree
• distribution shift
• concept stability