!! Jukan says these guys are DeepSeek's silicon partner.
!!Jukan says these guys are DeepSeek's silicon partner.Jukan @ ICML: @teortaxesTex verisilicon
Latest
By the by, I'm afraid this isn't great news for Western bros. As I've said before, the optimal situation for China is when they can have an open shari…
By the by, I'm afraid this isn't great news for Western bros. As I've said before, the optimal situation for China is when they can have…
insane copes if you don't like Nobel winners starting labs in China and believe it's a matter of opportunities, you can provide them with the same or …
insane copesif you don't like Nobel winners starting labs in China and believe it's a matter of opportunities, you can provide them with the same or…
The hidden caste system of ML research…
The hidden caste system of ML research…Rosinality: There is a strict hierarchy of ML research.OptimizersObjectivesArchitecturesDataEven among data works, there is a hierarchy:Synthetic dataData mixturesFilteringPreprocessingAdding new data…
The chutzpah of being a R*manian and talking how Chyna steals more than it innovates. Not like France or Germany have much of a leg to stand on… but …
The chutzpah of being a R*manian and talking how Chyna steals more than it innovates. Not like France or Germany have much of a leg to…
Qwen RLVR that makes it a dry math-brained tool [in visible outputs]:
Qwen RLVR that makes it a dry math-brained tool [in visible outputs]:Wyatt Walls: Ask Qwen about its favorite sport and this where its mind immediately goes
Liang Wenfeng is an electronic engineer before he's an AI bro, software developer or a quant. That's why Frank Wang tried to recruit him to what becam…
Liang Wenfeng is an electronic engineer before he's an AI bro, software developer or a quant. That's why Frank Wang tried to recruit him to what…
> done right cheeky but this is exciting. Tencent is still using GQA in their best model, which I just criticized them for. Using landmark tokens is c…
> done rightcheekybut this is exciting. Tencent is still using GQA in their best model, which I just criticized them for. Using landmark tokens is clever.Rosinality:…
"now"
"now"Andrew Curran: DeepSeek are working on their own chip, it seems as though everyone is now.
Seemingly that's exactly what has happened I still trust westoids too much, alas one needs to have the same standards for "high trust society" journal…
Seemingly that's exactly what has happenedI still trust westoids too much, alasone needs to have the same standards for "high trust society" journalists as for Indian…