Metraipolis › archive › dc_aef2f4d5942f

$α$Transfer: Coefficient Transfer for Efficient Model Merging — arXiv:2610.07819

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.07819
License (as archived)
CC0-1.0
Archived
2026-10-07T11:23:44.441Z
SHA-256 of the archived text
15049bf8ca88f50d8a6ab8d5ec3d11f1e37635e8ce387e7e2cbf177ae15da4cb
Stable URL of this page
https://metraipolis.com/documento/dc_aef2f4d5942f

How to cite

$α$Transfer: Coefficient Transfer for Efficient Model Merging — arXiv:2610.07819 — arXiv (en), https://arxiv.org/abs/2610.07819 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_aef2f4d5942f (Metraipolis archive dc_aef2f4d5942f, sha256 15049bf8ca88f50d8a6ab8d5ec3d11f1e37635e8ce387e7e2cbf177ae15da4cb)
BibTeX
@misc{dcaef2f4d5942f,
  title = {$α$Transfer: Coefficient Transfer for Efficient Model Merging — arXiv:2610.07819},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.07819},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_aef2f4d5942f; sha256 15049bf8ca88f50d8a6ab8d5ec3d11f1e37635e8ce387e7e2cbf177ae15da4cb},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.07819. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Model merging offers a promising solution for combining multiple fine-tuned checkpoints into a single model through parameter arithmetic. However, finding optimal merging coefficients requires an extensive search that becomes prohibitively expensive as models scale in both size and number, due to high memory requirements and combinatorial growth in the search space. We show that, within the same model family, models exhibit highly congruent performance distributions over merging coefficients across different model sizes. This distributional similarity enables a practical paradigm we call \textit{$\alpha$Transfer}: searching for optimal coefficients on a small proxy model, then directly transfer them to larger target models. We verify $\alpha$Transfer across multiple merging methods, model families, and tasks. Experimental results demonstrate a 6$\times$ speedup and 70\% memory reduction on vision transformers, and a 20$\times$ speedup and 85\% memory reduction on large language models, while maintaining comparable performance. Our findings establish $\alpha$Transfer as an efficient and generalizable approach to scaling model merging.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.07819 — 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 2026-10-07T11:15:39.490Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).