Metraipolis › archive › dc_111e8fbda138

SinkPruner: Sink-Free Visual Token Pruning for Multimodal Large Language Models — arXiv:2609.01004

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.01004
License (as archived)
CC0-1.0
Archived
2026-10-17T06:33:50.903Z
SHA-256 of the archived text
c711ba09139951bc7fdf48ca509305513d4defb8eaed855343bd02c725da0464
Stable URL of this page
https://metraipolis.com/documento/dc_111e8fbda138

How to cite

SinkPruner: Sink-Free Visual Token Pruning for Multimodal Large Language Models — arXiv:2609.01004 — arXiv (en), https://arxiv.org/abs/2609.01004 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_111e8fbda138 (Metraipolis archive dc_111e8fbda138, sha256 c711ba09139951bc7fdf48ca509305513d4defb8eaed855343bd02c725da0464)
BibTeX
@misc{dc111e8fbda138,
  title = {SinkPruner: Sink-Free Visual Token Pruning for Multimodal Large Language Models — arXiv:2609.01004},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.01004},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_111e8fbda138; sha256 c711ba09139951bc7fdf48ca509305513d4defb8eaed855343bd02c725da0464},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.01004. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Despite their strong multimodal understanding ability, multimodal large language models (MLLMs) incur substantial computational overhead when processing long visual token sequences. To reduce inference costs, recent studies have explored visual token pruning through vision-centric or text-guided strategies. However, these methods often overlook high-norm outlier tokens, i.e., tokens with abnormally large feature norms, leading to suboptimal pruning decisions. In this work, we show that such high-norm outlier tokens are highly redundant in both feature and spatial dimensions, yet are often mistakenly preserved as informative cues by existing methods. Motivated by this observation, we propose SinkPruner, a training-free visual token pruning framework for efficient MLLM inference. SinkPruner follows a coarse-to-fine design with two key modules: a visual sanitizer that filters high-norm redundancies and alleviates attention sink and attention dispersion, and a text-guided pruner that further retains tokens semantically aligned with the text query. Extensive experiments on twelve image-language and four video-language benchmarks demonstrate the effectiveness, efficiency, and generalizability of our framework. Notably, SinkPruner preserves 96.5% (91.8%) of the original performance of LLaVA-1.5 (Qwen2.5-VL) under an 89% token reduction. Experiments further indicate that our visual sanitizer exhibits promising transferability in enhancing the performance of existing pruning methods. Our code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.01004 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T06:15:12.848Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).