caleb 🐮 @clbswrs

I shall be one of those who make things beautiful. art, ML, 🏀, @pollilabs clbswrs.co USA Joined April 2018

Tweets

1K
Followers

254
Following

2K
Likes

38K

caleb 🐮 @clbswrs

a month ago

first claude weekly limits now chatty quarterly upload limits 🥺

0 0 2 56 0

Download Image

caleb 🐮 @clbswrs

2 months ago

Robert Schneider was the best math instructor of my life AND gave me the chillest fun fact to sneak in when around cool-cat neutral milk hotel-knowers

Lock @evildarkempath

2 months ago

Robert Schneider was the best math instructor of my life AND gave me the chillest fun fact to sneak in when around cool-cat neutral milk hotel-knowers

1 0 27 9K 11

0 1 3 811 0

typedfemale @typedfemale

2 months ago

presenting: big jeff's trainium hell

85 187 2K 169K 623

Download Video

John David Pressman @jd_pressman

2 months ago

Me last night: "So what stands out to me about this model. Is that it doesn't do the thing language models normally do where they kind of avoid detail? Like, a human will write about things using specific names and places. And if you pay close attention to LLM writing they…

xlr8harder @xlr8harder

2 months ago

22 56 933 88K 303

Download Image

2 11 195 29K 62

Bernard Stanford ✡︎ @stanfordNYC

2 months ago

@johncoogan Hey John, this also puts an interesting spin on California's ban on noncompetes. Traditionally, noncompetes are a burden for rank and file employees. But it seems exec-level noncompetes could be good for ordinary workers, to prevent c-suite from jumping ship like this?

3 1 49 6K 3

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞) @teortaxesTex

2 months ago

@main_horse Yeah it's literally a V3/R1 with a slightly enlarged hidden dim, It matches in all other ways. Would be real crazy if… and explain why DeepSeek is not eager to adopt Huawei compute for training.

1 1 30 4K 3

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞) @teortaxesTex

2 months ago

Extremely bad if true, unprecedented on this level of actors involved: Huawei covertly upcycled Qwen 2.5-14B for their Pangu MoE, Mixtral-like. Rumors go darker than that, but this much can be empirically investigated. @nopainkiller anything to say?