paper-with-me

Papers

Self-Expansion of Pre-trained Models with Mixture of Adapters for Continual Learning

2024-03-27 · CVPR 2025 1 · Huiyi Wang, Haodong Lu, Lina Yao, Dong Gong

Continual learning (CL) aims to continually accumulate knowledge from a non-stationary data stream without catastrophic forgetting of learned knowledge, requiring a balance between stability and adaptability. Relying on the generalizable representation in pre-trained models (PTMs), PTM-based CL methods perform effective continual adaptation on downstream tasks by adding learnable adapters or prompts upon the frozen PTMs. However, many existing PTM-based CL methods use restricted adaptation on a fixed set of these modules to avoid forgetting, suffering from limited CL ability. Periodically adding task-specific modules results in linear model growth rate and impaired knowledge reuse. We propose Self-Expansion of pre-trained models with Modularized Adaptation (SEMA), a novel approach to enhance the control of stability-plasticity balance in PTM-based CL. SEMA automatically decides to reuse or add adapter modules on demand in CL, depending on whether significant distribution shift that cannot be handled is detected at different representation levels. We design modular adapter consisting of a functional adapter and a representation descriptor. The representation descriptors are trained as a distribution shift indicator and used to trigger self-expansion signals. For better composing the adapters, an expandable weighting router is learned jointly for mixture of adapter outputs. SEMA enables better knowledge reuse and sub-linear expansion rate. Extensive experiments demonstrate the effectiveness of the proposed self-expansion method, achieving state-of-the-art performance compared to PTM-based CL methods without memory rehearsal. Code is available at https://github.com/huiyiwang01/SEMA-CL.

📄 PDF Abstract BibTeX arXiv:2403.18886

Code (1)

huiyiwang01/sema-cl 공식 구현 pytorch

Tasks

Continual Learning

Methods 이 논문이 사용한 방법론

SET Dynamic Sparse Training method where weight mask is updated randomly periodically
Focus 설명 없음
Adapter 설명 없음

Similar Papers 제목 키워드 기반

Boosting Continual Learning of Vision-Language Models via Mixture-of-Experts Adapters

2024-03-18 · CVPR 2024 1 · Jiazuo Yu, Yunzhi Zhuge, Lu Zhang, Ping Hu 외

Continual learning can empower vision-language models to continuously acquire new knowledge, without the need for access to the entire historical dataset. However, mitigating the performance degradation in large-scale mo…

Continual LearningIncremental LearningLifelong learningMixture-of-Experts+1

MILE: Mixture of Incremental LoRA Experts for Continual Semantic Segmentation across Domains and Modalities

2026-05-05 · Shishir Muralidhara, Didier Stricker, René Schuster arxiv

Continual semantic segmentation requires models to adapt to new domains or modalities without sacrificing performance on previously learned tasks. Expert-based learning, in which task-specific modules specialize in diffe…

Continual Semantic Segmentation

DIMoE-Adapters: Dynamic Expert Evolution for Continual Learning in Vision-Language Models

2026-05-08 · Mengxin Qin, Xiang Zhang, Xi Wang, Kun Wei 외 arxiv

Continual learning enables vision-language models to accumulate knowledge and adapt to evolving tasks without retraining from scratch. However, in multi-domain task-incremental learning, large domain shifts intensify the…

Incremental LearningContinual Learning

LLMs Can Evolve Continually on Modality for X-Modal Reasoning

2024-10-26 · Jiazuo Yu, Haomiao Xiong, Lu Zhang, Haiwen Diao 외

Multimodal Large Language Models (MLLMs) have gained significant attention due to their impressive capabilities in multimodal understanding. However, existing methods rely heavily on extensive modal-specific pretraining …

Continual Learningmultimodal interaction

Little By Little: Continual Learning via Self-Activated Sparse Mixture-of-Rank Adaptive Learning

2025-06-26 · Haodong Lu, Chongyang Zhao, Jason Xue, Lina Yao 외

Continual learning (CL) with large pre-trained models is challenged by catastrophic forgetting and task interference. Existing LoRA-based Mixture-of-Experts (MoE) approaches mitigate forgetting by assigning and freezing …

Continual LearningMixture-of-Experts