VertiKV: Vertical-Integrity KV Cache Compression for Efficient Multimodal Long-Context Inference
Open Access
- 9 November 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 12367-12376
Abstract
No abstract availableFunding Information
- the Key Research and Development Program of Zhejiang Province (No. 2025C01026)
- National Natural Science Foundation of China (No. 62402429, U24A20326, 62441236)
- Ningbo Yongjiang Talent Introduction Programme (2023A-397-G)
- Young Elite Scientists Sponsorship Program by CAST (2024QNRC001)
This publication has 24 references indexed in Scilit:
- ZipVL: Accelerating Vision-Language Models Through Dynamic Token Sparsity2025
- iEdit: Localised Text-guided Image Editing with Weak Supervision2024
- MMMU: A Massive Multi-Discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI2024
- Structured Pruning for Deep Convolutional Neural Networks: A SurveyIEEE Transactions on Pattern Analysis and Machine Intelligence, 2023
- DocVQA: A Dataset for VQA on Document Images2021
- TextCaps: A Dataset for Image Captioning with Reading Comprehension2020
- ALFRED: A Benchmark for Interpreting Grounded Instructions for Everyday Tasks2020
- Robust Change Captioning2019
- Towards VQA Models That Can Read2019
- Flickr30k Entities: Collecting Region-to-Phrase Correspondences for Richer Image-to-Sentence Models2015