Motif 3: Technical Report

Motif 3: Technical Report

Motif 3 is a large sparse mixture-of-experts language model using grouped differential latent attention and specialized training techniques to achieve strong reasoning, coding, and…

Hugging Face · Daily Papers ·Junghwan Lim, Joon Son Chung · ·▲ 24 upvotes

Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.

Autores: Junghwan Lim, Joon Son Chung, Sungmin Lee, Wai Ting Cheung, Gihun Cho, Minsu Ha

  • 24 upvotes da comunidade
  • Temas: Mixture-of-Experts, sparse MoE, routed experts, Grouped Differential Latent Attention, Multi-head Latent Attention, manifold-constrained hyper-connections

Resumo

Resumo original (em inglês), extraído do paper:

Motif 3 is a large sparse mixture-of-experts language model using grouped differential latent attention and specialized training techniques to achieve strong reasoning, coding, and long-context performance.

Onde ler

compartilhar: