Yes, Deepseek introduced this optimisation of a common base "expert" that's alwa... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

phire on April 6, 2025 | parent | context | favorite | on: The Llama 4 herd

Yes, Deepseek introduced this optimisation of a common base "expert" that's always loaded. Llama 4 uses it too.

regularfry on April 7, 2025 [–]

I had a sneaking suspicion that I wouldn't be the first to think of it.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact