ODE-M formulates continual model merging as a barrier-aware ODE trajectory in parameter space, using first-order feedback and a utility-aware schedule to balance retained knowledge and new task performance.
Model merging and safety alignment: One bad model spoils the bunch
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
fields
cs.LG 3roles
background 2polarities
background 2representative citing papers
ADR achieves theoretically zero-forgetting class-incremental graph learning by combining backpropagation adaptation with ridge-regression-based layer-wise merging of GNN linear transformations.
The paper introduces a new taxonomy for model merging methods and reviews their applications in LLMs, MLLMs, continual learning, multi-task learning, and other subfields while outlining open challenges.
citing papers explorer
-
Unlocking the Potential of Continual Model Merging: An ODE Perspective
ODE-M formulates continual model merging as a barrier-aware ODE trajectory in parameter space, using first-order feedback and a utility-aware schedule to balance retained knowledge and new task performance.
-
Analytic Drift Resister for Non-Exemplar Continual Graph Learning
ADR achieves theoretically zero-forgetting class-incremental graph learning by combining backpropagation adaptation with ridge-regression-based layer-wise merging of GNN linear transformations.
-
Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunities
The paper introduces a new taxonomy for model merging methods and reviews their applications in LLMs, MLLMs, continual learning, multi-task learning, and other subfields while outlining open challenges.