Moment-KV: Momentum-Based Decode-Time KV Cache Compression for Long Generation 事件

Moment-KV: Momentum-Based Decode-Time KV Cache Compression for Long Generation arXiv:2605.29873v1 Announce Type: new Abstract: Key-Value (KV) cache remains a major bottleneck for deploying Large Language Models (LLMs) in long-generation tasks. Prior work often applies uniform compression across both prefill and decoding caches, but compressing the prefill cache degrades performance by corrupting critical context. While preserving the prefill cache is essential, decoding-phase compression remain