Research map / Continual learning / Online and replay-based
Continual Model Merging without Data: Dual Projections for Balancing Stability and Plasticity
Enneng Yang, Anke Tang, Li Shen, Guibing Guo, Xingwei Wang, Xiaochun Cao, Jie Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes a data-free continual model merging method using dual projections to balance stability and plasticity when expert models arrive sequentially.
Problem
Most model merging methods assume all expert models are available simultaneously, which is impractical when models arrive sequentially; existing continual model merging approaches must balance stability (retaining earlier models' knowledge) and plasticity (integrating new models).
Approach
Uses dual projections to merge sequentially arriving models without access to training data, balancing stability and plasticity in the merged model.
Closest papers by meaning
- BECAME: Bayesian Continual Learning with Adaptive Model Merging (ICML 2025 · similarity 0.92)
- Merging on the Fly Without Retraining: A Sequential Approach to Scalable Continual Model Merging (NeurIPS 2025 · similarity 0.91)
- Train with Perturbation, Infer after Merging: A Two-Stage Framework for Continual Learning (NeurIPS 2025 · similarity 0.90)
- Modeling Multi-Task Model Merging as Adaptive Projective Gradient Descent (ICML 2025 · similarity 0.89)
- Mixture-Trained Merging for Unified Multi-Objective Models (NeurIPS 2026 · similarity 0.88)
- MergeBench: A Benchmark for Merging Domain-Specialized LLMs (NeurIPS 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
