ReForge: Refining Merged Models with Anchor-Regularized Regression

Excerpt

arXiv:2605.12843v2 Announce Type: replace Abstract: Model merging aims to combine multiple task-specific expert models into a single model without joint retraining, offering a practical alternative to multi-task learning when data access or computational budget is limited. Existing model merging methods rarely exploit strong merged models as priors for further improvement. To address this limitation, we propose ReForge, a bilevel optimization framework that formulates module-wise refinement as B