AIが既存性能を維持したまま新規タスクを学習可能に（Researchers Improve AI’s Ability to Learn New Tasks Without Sacrificing Performance）

2026-05-19 ノースカロライナ州立大学（NC State）

Image credit: Luke Jones.

＜関連情報＞

• https://news.ncsu.edu/2026/05/ai-continual-learning-cheem/
• https://arxiv.org/abs/2303.08250

CHEEM：再利用、新規開発、適応、スキップによる継続的学習 ― 階層的な探索・活用アプローチ CHEEM: Continual Learning by Reuse, New, Adapt and Skip — A Hierarchical Exploration-Exploitation Approach

Chinmay Savadikar, Michelle Dai, Tianfu Wu
arXiv  last revised 1 Apr 2026 (this version, v5)
DOI:https://doi.org/10.48550/arXiv.2303.08250

Abstract

To effectively manage the complexities of real-world dynamic environments, continual learning must incrementally acquire, update, and accumulate knowledge from a stream of tasks of different nature without suffering from catastrophic forgetting of prior knowledge. While this capability is innate to human cognition, it remains a significant challenge for modern deep learning systems. At the heart of this challenge lies the stability-plasticity dilemma: the need to balance leveraging prior knowledge, integrating novel information, and allocating model capacity adaptively based on task complexity and synergy. In this paper, we propose a novel exemplar-free class-incremental continual learning (ExfCCL) framework that addresses these issues through a Hierarchical Exploration-Exploitation (HEE) approach. The core of our method is a HEE-guided efficient neural architecture search (HEE-NAS) that enables a learning-to-adapt backbone via four primitive operations – reuse, new, adapt, and skip – thereby serving as an internal memory that dynamically updates selected components across streaming tasks. To address the task ID inference problem in ExfCCL, we exploit an external memory of task centroids proposed in the prior art. We term our method CHEEM (Continual Hierarchical-Exploration-Exploitation Memory). CHEEM is evaluated on the challenging MTIL and VDD benchmarks using both Tiny and Base Vision Transformers and a proposed holistic Figure-of-Merit (FoM) metric. It significantly outperforms state-of-the-art prompting-based continual learning methods, closely approaching full fine-tuning upper bounds. Furthermore, it learns adaptive model structures tailored to individual tasks in a semantically meaningful way. Our code is available at this https URL .