VLCache: Computing 2% Vision Tokens and Reusing 98% for Vision-Language Inference

v1v2 (latest)

VLCache: Computing 2% Vision Tokens and Reusing 98% for Vision-Language Inference

15 December 2025

Shengling Qin

Hao Yu

Chenxin Wu

Zheng Li

Yizhong Cao

Zhengyang Zhuge

Yuxin Zhou

Wentao Yao

Yi Zhang

Zhengheng Wang

Shuai Bai

Jianwei Zhang

Junyang Lin

ArXiv (abs)PDF HTML

Papers citing "VLCache: Computing 2% Vision Tokens and Reusing 98% for Vision-Language Inference"

0 / 0 papers shown

Title
No papers found