#natural language process...
Jul 2026
Mixture-of-Translators: Translating KV Caches Across Heterogeneous Large Language Models
Mixture-of-Translators (MoT), a cache translation framework that maps context KV caches from a source LLM into the cache space of a target LLM, is proposed, demonstrating scalable KV cache reuse across heterogeneous LLMs.
Jin-woo Lee, Minkyung Song, Junghyun Oh et al.
· arXiv.org · 4 citations