{"ID":6138098,"CreatedAt":"2026-07-09T01:07:32.349475501Z","UpdatedAt":"2026-07-11T06:05:24.908464437Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.07033","arxiv_id":"2607.07033","title":"AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning","abstract":"Large vision-language models incur substantial inference costs because high-resolution inputs introduce thousands of visual tokens, many of which are redundant for a given query. Existing pruning methods often combine query relevance and token diversity, yet these objectives can conflict under aggressive compression: relevance-driven selection may overconcentrate the budget on correlated local evidence, while diversity-driven selection may suppress indispensable tokens or retain distinct but uninformative regions. We introduce AnchorPrune, a training-free framework that first constructs a protected relevance anchor and then expands it with complementary visual context. AnchorPrune adaptively determines the anchor size from the novelty profile of relevance-ranked tokens, preserving a compact set of query-critical evidence, and allocates the remaining budget through importance-weighted novelty to recover informative, non-redundant context relative to the anchor. This ordered design prevents contextual expansion from displacing indispensable query cues while improving overall visual coverage. AnchorPrune is lightweight, architecture-aware, and requires neither retraining nor model modification. Across image and video vision-language models and benchmarks, it consistently improves the accuracy-efficiency trade-off over training-free baselines, particularly under severe compression. On LLaVA-NeXT-7B, AnchorPrune preserves 97.6% of full-token performance using only 160 of 2,880 visual tokens. These results establish relevance-anchored contextual expansion as an effective principle for efficient multimodal inference. Code is available at https://github.com/MULTI-cau/AnchorPrune.","short_abstract":"Large vision-language models incur substantial inference costs because high-resolution inputs introduce thousands of visual tokens, many of which are redundant for a given query. Existing pruning methods often combine query relevance and token diversity, yet these objectives can conflict under aggressive compression: r...","url_abs":"https://arxiv.org/abs/2607.07033","url_pdf":"https://arxiv.org/pdf/2607.07033v1","authors":"[\"Kyuan Oh\",\"Bumsoo Kim\"]","published":"2026-07-08T06:07:06Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.AI\"]","methods":"[\"Language Model\"]","has_code":false,"code_links":[{"ID":614050,"CreatedAt":"2026-07-09T01:07:32.349475501Z","UpdatedAt":"2026-07-09T01:07:32.349475501Z","DeletedAt":null,"paper_id":6138098,"paper_url":"https://arxiv.org/abs/2607.07033","paper_title":"AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning","repo_url":"https://github.com/MULTI-cau/AnchorPrune","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}