Separating Common and Unique Directions for Model Merging
cs.LG, cs.AI
Submitted: 2026-11-03
Updated: 2026-11-04
Comments: Accepted by NeurIPS 2026
Code: https://github.com/stwang666/CUSD
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Model Merging in the Era of Large Language Models: Methods, Applications, and Future Directions
- Model Merging Scaling Laws in Large Language Models
- Model Merging via Data-Free Covariance Estimation
- ACE-Merging: Data-Free Model Merging with Adaptive Covariance Estimation
- A Practitioner's Guide to Continual Multimodal Pretraining
- What Matters for Model Merging at Scale?
- Editing Models with Task Arithmetic
- Localizing Task Information for Improved Model Merging and Compression
- No Task Left Behind: Isotropic Model Merging with Common and Task-Specific Subspaces
- The Llama 3 Herd of Models
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Subspace-Boosted Model Merging
- Model merging with SVD to tie the Knots
- Multi-Task Model Merging via Adaptive Weight Disentanglement
- Modeling Multi-Task Model Merging as Adaptive Projective Gradient Descent
- Whoever Started the Interference Should End It: Guiding Data-Free Model Merging via Task Vectors
- Dataless Knowledge Fusion by Merging Weights of Language Models
- ZipIt! Merging Models from Different Tasks without Training
- AdaMerging: Adaptive Model Merging for Multi-Task Learning
- RoBERTa: A Robustly Optimized BERT Pretraining Approach
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis