Robustness and Regularization in Hierarchical Re-Basin

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Franke, Benedikt, Heinrich, Florian, Lange, Markus, Raulf, Arne
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916026576273408
author Franke, Benedikt
Heinrich, Florian
Lange, Markus
Raulf, Arne
author_facet Franke, Benedikt
Heinrich, Florian
Lange, Markus
Raulf, Arne
contents This paper takes a closer look at Git Re-Basin, an interesting new approach to merge trained models. We propose a hierarchical model merging scheme that significantly outperforms the standard MergeMany algorithm. With our new algorithm, we find that Re-Basin induces adversarial and perturbation robustness into the merged models, with the effect becoming stronger the more models participate in the hierarchical merging scheme. However, in our experiments Re-Basin induces a much bigger performance drop than reported by the original authors.
format Preprint
id arxiv_https___arxiv_org_abs_2510_09174
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Robustness and Regularization in Hierarchical Re-Basin
Franke, Benedikt
Heinrich, Florian
Lange, Markus
Raulf, Arne
Machine Learning
This paper takes a closer look at Git Re-Basin, an interesting new approach to merge trained models. We propose a hierarchical model merging scheme that significantly outperforms the standard MergeMany algorithm. With our new algorithm, we find that Re-Basin induces adversarial and perturbation robustness into the merged models, with the effect becoming stronger the more models participate in the hierarchical merging scheme. However, in our experiments Re-Basin induces a much bigger performance drop than reported by the original authors.
title Robustness and Regularization in Hierarchical Re-Basin
topic Machine Learning
url https://arxiv.org/abs/2510.09174