1
Moment-KV: Momentum-Based Decode-Time KV Cache Compression for Long Generation
长文本生成中的KV缓存压缩新方法，利用动量策略在解码时高效压缩，大幅降低内存占用。
arXiv:2605.29873v1 Announce Type: new Abstract: Key-Value (KV) cache remains a major bottleneck for deploying Large Language Models (LLMs) in long-gen…