Centering before Pruning: Lightweight Geometry Correction for Diversity-Based Visual Token Pruning in LVLMs

arXiv:2608.30263 · cs.CV, cs.AI · Submitted 2026-08-31 · Read on arXiv

cs.CV, cs.AI

Submitted: 2026-08-31

Updated: 2026-08-31

Code: https://github.com/EvolvingLMMs-Lab/lmms-eval

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Pruning — 59× in this paper · explained in 4 episodes
• Benchmark — 32× in this paper · explained in 2 episodes
• Cosine Similarity — 26× in this paper · explained in 3 episodes
• Ablation — 10× in this paper · explained in 1 episode
• Redundancy — 8× in this paper · explained in 2 episodes
• Ablation Studies — 7× in this paper · explained in 1 episode
• Computational overhead — 6× in this paper · explained in 2 episodes
• Hallucination — 5× in this paper · explained in 3 episodes
• Conference on Computer Vision and Pattern Recognition — 4× in this paper
• Instruction Tuning — 4× in this paper · explained in 2 episodes
• Accuracy — 3× in this paper · explained in 1 episode
• Temporal Reasoning — 2× in this paper · explained in 2 episodes
• Transformer — 2× in this paper · explained in 2 episodes
• Ablation study — 1× in this paper · explained in 1 episode
• Adaptation — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 3 episodes
• Entanglement — 1× in this paper · explained in 4 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Homogeneity — 1× in this paper · explained in 2 episodes
• Multimodal Reasoning — 1× in this paper · explained in 5 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Semantic Similarity — 1× in this paper · explained in 2 episodes
• Tokenization — 1× in this paper · explained in 2 episodes

Sources

• Qwen2.5-VL Technical Report
• LLaVA-OneVision: Easy Visual Task Transfer
• InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models