Skip to content
AI News HubLIVE
Original source2 min read

Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization

Summary

CoverPruner is a training-free visual token pruning method for vision-language models. Rather than only deciding which tokens to keep, it asks which surviving token should represent each pruned token, formulating pruning as query-weighted representational coverage maximization. Using projector-space coverage and a lightweight first-layer attention probe, CoverPruner achieves the best average accuracy across multiple VLM architectures and compression rates, with the largest gains typically appearing under aggressive compression.

SourcearXiv Computer VisionAuthor: Qingchan Zhu, Weihang You, Hanqi Jiang, Changdi Yang, Tianming Liu, Geng Yuan
Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 2 Sep 2026]

Title:Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization

View a PDF of the paper titled Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization, by Qingchan Zhu and 5 other authors

View PDF HTML (experimental)

Abstract:Visual token pruning reduces the inference cost of vision-language models (VLMs), but most methods only ask which tokens to keep. This retained-token view can keep redundant high-scoring tokens while leaving discarded evidence without a close representative. We propose CoverPruner, a training-free pruner that asks the complementary demand-side question: after a token is removed, which surviving original token represents it for the target VLM? CoverPruner formulates pruning as Representational Coverage Maximization (RCM), covering the full projected visual-token set with query-weighted demand. It instantiates RCM with projector-space coverage and a lightweight first-layer attention probe. Across multiple VLM architectures and compression rates, CoverPruner achieves the best average accuracy among all compared methods, with the largest gains usually appearing under aggressive compression.

Comments: Accepted to EMNLP 2026 main

Subjects:

Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Machine Learning (cs.LG)

Cite as: arXiv:2609.03158 [cs.CV]

(or arXiv:2609.03158v1 [cs.CV] for this version)

https://doi.org/10.48550/arXiv.2609.03158

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Weihang You [view email] [v1] Wed, 2 Sep 2026 20:51:02 UTC (5,908 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization, by Qingchan Zhu and 5 other authors

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.CV

new | recent | 2026-09

Change to browse by:

cs cs.CL cs.LG

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • Asks which surviving token represents each pruned token, complementing keep-only strategies.
  • Formulates token pruning as query-weighted Representational Coverage Maximization (RCM).
  • Uses projector-space coverage and a lightweight first-layer attention probe, requiring no training.
  • Achieves the best average accuracy across VLM architectures and compression rates, especially with aggressive compression.

Highlights and analysis are generated automatically and may contain errors. Check the original source.