{"ID":5675285,"CreatedAt":"2026-07-03T01:40:09.565152011Z","UpdatedAt":"2026-07-07T01:06:03.009715918Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.01978","arxiv_id":"2607.01978","title":"Multimodal Knowledge Edit-Scoped Generalization for Online Recursive MLLM Editing","abstract":"Online multimodal knowledge editing requires injecting a continual stream of visual-textual corrections into multimodal large language models (MLLMs) with bounded overhead and minimal disruption to unrelated behaviors. Existing editors mainly emphasize edit reliability and long-horizon stability, but rarely control the semantic boundary of each edit. Our pilot analyses of post-edit behaviors and internal neuronal activities reveal a scope gap behind reliable edits: instance-level success neither guarantees transfer to valid cross-modal variants nor prevents leakage to unrelated inputs, while edit-related cross-modal responses concentrate in deeper semantic layers. Therefore, we formulate Edit-Scoped Generalization, reframing online MLLM editing from merely correcting an instance to controlling the propagation boundary of each edit. To this end, we propose ScopeEdit, a scope-aware online editor that decomposes each update into a modality-local absorption branch and an evidence-gated shared generalization branch. The local branch supports stable edit absorption, whereas the shared branch enables cross-modal propagation only when visual and textual evidence are sufficiently aligned. Both branches perform scope-separated write geometries in orthogonal low-rank spaces and maintain branch-wise preconditioners via Sherman--Morrison recursions, yielding constant per-edit overhead. Extensive experiments across diverse benchmarks, long-horizon edit streams, MLLM backbones, real-world VLKEB scenarios, and complex vision-language architectures show that ScopeEdit consistently improves the trade-off between in-scope cross-modal transfer and out-of-scope locality, while preserving edit reliability, stability and online efficiency. Our code is available at https://github.com/lab-klc/ScopeEdit.","short_abstract":"Online multimodal knowledge editing requires injecting a continual stream of visual-textual corrections into multimodal large language models (MLLMs) with bounded overhead and minimal disruption to unrelated behaviors. Existing editors mainly emphasize edit reliability and long-horizon stability, but rarely control the...","url_abs":"https://arxiv.org/abs/2607.01978","url_pdf":"https://arxiv.org/pdf/2607.01978v1","authors":"[\"Siyuan Li\",\"Youyuan Zhang\",\"Ruitong Liu\",\"Junxi Wang\",\"Jing Li\"]","published":"2026-07-02T10:10:19Z","proceeding":"cs.AI","tasks":"[\"cs.AI\",\"cs.CL\",\"cs.CV\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false,"code_links":[{"ID":613894,"CreatedAt":"2026-07-03T01:40:09.565152011Z","UpdatedAt":"2026-07-03T01:40:09.565152011Z","DeletedAt":null,"paper_id":5675285,"paper_url":"https://arxiv.org/abs/2607.01978","paper_title":"Multimodal Knowledge Edit-Scoped Generalization for Online Recursive MLLM Editing","repo_url":"https://github.com/lab-klc/ScopeEdit","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}