News for hackers. Stuff that matters.

A comment on Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch. Parent comment

Re:Tongyi DeepResearch – open-source 30B MoE Model that riva... by pama on Sunday November 02, 2025 @08:35PM (#45795071)
> only uses 1/18th of the total parameters per-query.

only uses 1/18th of the total parameters per token. It may use the large fraction of them in a single query.

[ Parent | on HN ]
Re:Tongyi DeepResearch – open-source 30B MoE Model that riva... by ainch on Monday November 03, 2025 @02:58PM (#45803670)
That's a good correction, thanks.
[ Parent | on HN ]