一作ACL 2026Findings· Findings of ACL
MM - S hift KV : Decode-Aware Prefill-Stage KV Selection for Multimodal Large Language Models
Jinsong Shu · Chenyang Wu · Zhongle Xie · Baokun Wang · Lidan Shou
摘要
Key-Value (KV) caching is essential for efficient inference in multimodal large language models (MLLMs), yet its memory footprint grows linearly with context length and becomes a major bottleneck due to the large number of visual tokens. Recent prefill-stage KV selection methods estimate KV impor...