{"id":"2607.26596","title":"Decoupled Visual Processing: Efficient Multimodal Adaptation via Modality-Specific Transformer Substitution","url":"https://codeofpaper.com/paper/2607.26596","arxiv_url":"https://arxiv.org/abs/2607.26596","published":"2026-07-29","code":"none","best":null,"most_starred_community":null,"implementations":[],"possible_matches":0}