Xiaomi’s MiMo-V3 to adopt new architecture as HySparse2 cuts long-context costs
Xiaomi MiMo lead Fuli Luo said MiMo-V3 will adopt a new architecture. Its core technology, HySparse2, is designed to reduce the cost of long-context inference and improve retrieval for agentic workloads. At a context le…
以下正文同步自 动点科技 TechNode，版权归原站所有，已转换为易读排版。
Xiaomi MiMo lead Fuli Luo said MiMo-V3 will adopt a new architecture. Its core technology, HySparse2, is designed to reduce the cost of long-context inference and improve retrieval for agentic workloads. At a context length of 1 million tokens, HySparse2 reduces prefill computation by 5.02 times and cuts the KV cache by 4.5 times. Xiaomi also reported higher MRCRv2 and RULER-v2 scores, along with lower AgentPPL and LongPPL results. [arXiv]
正文由 探How 从来源站点 RSS 同步，内容未经改写；遇到排版缺失或需要图片、视频时请以原文为准。