cs.AINov 13, 2026

A Spectral Framework for Universal Proxy Regularization in Model Merging

Authors: Shangge Liu, Danyue Qian, Qi Fan, Yinghuan Shi, Wenbin Li, Lei Wang, Yang Gao, Dacheng Tao

Organizations: State Key Laboratory of Novel Software Technology, Nanjing University, China · School of Computing and Information Technology, University of Wollongong, Australia · Nanyang Technological University, Singapore

Abstract

Model merging constructs multi-task models without expensive joint training but often suffers from parameter interference. Some solutions mitigate this by constraining finetuning using data from the other tasks to be merged, violating the core independence and privacy assumption of model merging. Recent work offers preliminary empirical evidence that a universal, task-agnostic proxy distribution can substitute for such data, but leaves open when and why it suffices. To this end, we propose a Spectral Framework for Universal Proxy Regularization, which characterizes the relationship between a proxy distribution and the distribution of other tasks to be merged through spectral coverage (αα) and leakage (ββ). Based on the framework, we derive a unified parameter interference bound: smaller (α,β)(α, β) provably yields lower parameter interference through better spectral alignment. To exploit such a proxy, we augment the training objective with a proxy regularizer. Evaluating this regularizer during fine-tuning requires a concrete mechanism. We study three instantiations, including (J) an exact evaluation via Jacobian-vector products, (F) a frozen curvature approximation using the Fisher information matrix, and (O) an output-drift formulation that constrains the model's functional deviation on the proxy directly. We prove that all three coincide under the NTK linearization, where existing analyses of such regularizers stop, and go further to characterize standard non-linear fine-tuning, bounding how far each instantiation departs from its linearized reference and obtain a three-way hierarchy, zero for (J), proxy-independent for (F), and proxy-quality-scaling for (O). Empirical validation on the commonly used vision-language model CLIP confirms these theoretical predictions.

Figures & tables

Explore similar work

CardsList
  1. Closed-Form Spectral Regularization for Multi-Task Model Merging

    Jun 5, 2026Yongxian Wei, Runxi Cheng, Xingxuan Zhang +4Spectral RegularizationMulti-Task Learning

  2. A Broader Look at Model Merging: Rethinking Implicit Regularization Induced by Task Arithmetic

    Oct 6, 2026Sin-Han Yang, Shih-Cheng Huang, Chieh-Yen Lin +3Task Vector MergingMulti-Task Optimization

  3. Task Alignment: A Simple Proxy for Practical Model Merging Across Diverse Vision Tasks

    Apr 14, 2026Pau de Jorge, César Roberto de Souza, Björn Michele +5Model Merging