visual-token-pruning

Official code for paper: [CLS] Attention is All You Need for Training-Free Visual Token Pruning: Make VLM Inference Faster.

[NeurIPS 2025] Official code for paper: Beyond Attention or Similarity: Maximizing Conditional Diversity for Token Pruning in MLLMs.

[TCSVT] Official repository of the paper "A Glimpse to Compress: Dynamic Visual Token Pruning for Large Vision-Language Models"

[ICCV 2025] Official code for paper: Beyond Text-Visual Attention: Exploiting Visual Cues for Effective Token Pruning in VLMs

[ICLR 2026] AgilePruner: An Empirical Study of Attention and Diversity for Adaptive Visual Token Pruning in Large Vision-Language Models

this is the official code repo of our work: Beyond Intermediate States: Explaining Visual Redundancy through Language.https://arxiv.org/abs/2503.20540

Code and reproducibility artifacts for auditing spatial provenance in visual token pruning.

Official LLaVA implementation of STAR-Pro (arXiv:2609.05916): training-free visual token pruning with pivoted QR and progressive refinement.

Reliability-constrained visual-token budgeting for energy-efficient vision-language model inference.