{"ID":2840095,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.14569","arxiv_id":"2511.14569","title":"Task Addition and Weight Disentanglement in Closed-Vocabulary Models","abstract":"Task arithmetic has recently emerged as a promising method for editing pre-trained \\textit{open-vocabulary} models, offering a cost-effective alternative to standard multi-task fine-tuning. However, despite the abundance of \\textit{closed-vocabulary} models that are not pre-trained with language supervision, applying task arithmetic to these models remains unexplored. In this paper, we deploy and study task addition in closed-vocabulary image classification models. We consider different pre-training schemes and find that \\textit{weight disentanglement} -- the property enabling task arithmetic -- is a general consequence of pre-training, as it appears in different pre-trained closed-vocabulary models. In fact, we find that pre-trained closed-vocabulary vision transformers can also be edited with task arithmetic, achieving high task addition performance and enabling the efficient deployment of multi-task models. Finally, we demonstrate that simple linear probing is a competitive baseline to task addition. Overall, our findings expand the applicability of task arithmetic to a broader class of pre-trained models and open the way for more efficient use of pre-trained models in diverse settings.","short_abstract":"Task arithmetic has recently emerged as a promising method for editing pre-trained \\textit{open-vocabulary} models, offering a cost-effective alternative to standard multi-task fine-tuning. However, despite the abundance of \\textit{closed-vocabulary} models that are not pre-trained with language supervision, applying t...","url_abs":"https://arxiv.org/abs/2511.14569","url_pdf":"https://arxiv.org/pdf/2511.14569v1","authors":"[\"Adam Hazimeh\",\"Alessandro Favero\",\"Pascal Frossard\"]","published":"2025-11-18T15:12:21Z","proceeding":"cs.LG","tasks":"[\"cs.LG\"]","methods":"[\"Vision Transformer\",\"Transformer\"]","has_code":false}