paper-with-me

홈 › Papers

Continual Learning in Vision-Language Models via Aligned Model Merging

2025-05-30 · Ghada Sokar, Gintare Karolina Dziugaite, Anurag Arnab, Ahmet Iscen, Pablo Samuel Castro, Cordelia Schmid

Continual learning is conventionally tackled through sequential fine-tuning, a process that, while enabling adaptation, inherently favors plasticity over the stability needed to retain prior knowledge. While existing approaches attempt to mitigate catastrophic forgetting, a bias towards recent tasks persists as they build upon this sequential nature. In this work we present a new perspective based on model merging to maintain stability while still retaining plasticity. Rather than just sequentially updating the model weights, we propose merging newly trained task parameters with previously learned ones, promoting a better balance. To maximize the effectiveness of the merging process, we propose a simple mechanism that promotes learning aligned weights with previous ones, thereby avoiding interference when merging. We evaluate this approach on large Vision-Language Models (VLMs), and demonstrate its effectiveness in reducing forgetting, increasing robustness to various task orders and similarities, and improving generalization.

📄 PDF Abstract BibTeX arXiv:2506.03189

Code (0)

등록된 구현이 없습니다.

Tasks

Continual Learning

Similar Papers 제목 키워드 기반

MergeSlide: Continual Model Merging and Task-to-Class Prompt-Aligned Inference for Lifelong Learning on Whole Slide Images

2025-11-17 · Doanh C. Bui, Ba Hung Ngo, Hoai Luan Pham, Khang Nguyen 외 arxiv

Lifelong learning on Whole Slide Images (WSIs) aims to train or fine-tune a unified model sequentially on cancer-related tasks, reducing the resources and effort required for data transfer and processing, especially give…

class-incremental learningContinual Learning

Continual Vision-Language Learning for Remote Sensing: Benchmarking and Analysis

2026-04-01 · Xingxing Weng, Ruifeng Ni, Chao Pang, XiangYu Hao 외 arxiv

Current remote sensing vision-language models (RS VLMs) demonstrate impressive performance in image interpretation but rely on static training data, limiting their ability to accommodate continuously emerging sensing mod…

Continual Learning

MergeSurv: Merging-Based Continual Learning for Survival Analysis on Whole-Slide Images

2026-07-06 · Vu Minh Tran, Doanh C. Bui, Maï K. Nguyen, Khang Nguyen arxiv

Survival analysis on Whole Slide Images (WSIs) is important in computational pathology for prognosis estimation and treatment planning. However, existing survival models are typically trained independently for each cance…

Continual Learning

Forgetting of task-specific knowledge in model merging-based continual learning

2025-07-31 · Timm Hess, Gido M van de Ven, Tinne Tuytelaars arxiv

This paper investigates the linear merging of models in the context of continual learning (CL). Using controlled visual cues in computer vision experiments, we demonstrate that merging largely preserves or enhances share…

Continual Learning

Multi-Level Safety Continual Projection for Fine-Tuned Large Language Models without Retraining

2025-08-08 · Bing Han, Feifei Zhao, Dongcheng Zhao, Guobin Shen 외 arxiv

While fine-tuning services drive the rapid expansion of task capabilities in large language models (LLMs), they are often accompanied by the degradation and reorganization of safety-aligned representations, making models…