Metraipolis › archive › dc_eb810f0a74c4
Separating Common and Unique Directions for Data-Free Model Merging — arXiv:2611.04108
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.04108
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-04T12:29:41.744Z
- SHA-256 of the archived text
309425392b514806c5ad21a527e5b9e2809dd4a41b6b1bbde563af138f411956- Stable URL of this page
https://metraipolis.com/documento/dc_eb810f0a74c4
How to cite
Separating Common and Unique Directions for Data-Free Model Merging — arXiv:2611.04108 — arXiv (en), https://arxiv.org/abs/2611.04108 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eb810f0a74c4 (Metraipolis archive dc_eb810f0a74c4, sha256 309425392b514806c5ad21a527e5b9e2809dd4a41b6b1bbde563af138f411956)
BibTeX
@misc{dceb810f0a74c4,
title = {Separating Common and Unique Directions for Data-Free Model Merging — arXiv:2611.04108},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.04108},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eb810f0a74c4; sha256 309425392b514806c5ad21a527e5b9e2809dd4a41b6b1bbde563af138f411956},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04108. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Model merging combines multiple fine-tuned models without additional training, and arithmetic methods have emerged as a simple yet effective way to merge models without data. However, existing arithmetic methods treat task vectors in parameter space and apply a single element-wise rule to directions that can differ in mergeability. In a shared directional basis, we find a consistent two-regime structure: a low-dispersion common bulk shared across tasks and a high-dispersion task-dominated tail that carries disproportionate energy. These findings motivate Common--Unique Subspace Decomposition (CUSD), a data-free pre-merging method that separates task vectors into common and unique directional components, merges the common components with downstream arithmetic operator, and retains unique residuals separately. Experiments across $11$ base models, four model families, and four arithmetic-based merging operators show that CUSD improves 42 of 44 model--operator combinations and achieves up to 22.5% improvement over the corresponding baselines. Our code is publicly available at this https URL Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.04108 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-04T12:26:47.247Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).