Keep on to blur preview images; turn off to show them clearly
yes they have published a paper, because of course they have. I am always impressed by how simply and seriously they look at the field. It's not reactive hill-climbing busywork. «How do we create a generation-verification gap». «How do we fully remove human labeling». Then do it.
We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1
For all the people in the replies who have spent less than 30 minutes in a gym 😘 https://t.co/wUxhJbLSqL
Creator of @tailwindcss. Listener of Slayer. Austin 3:16.
Amazon grocery delivery is such a game changer! I scheduled days ahead to have the ingredients delivered right at my door. No need to go through the chaos of Thanksgiving grocery shopping 🛒
“The woman scooping Silicon Valley” — BBC・hacker turned builder, blogger & consultant・ex: Threads, Instagram, etc
i have seen many people try this and go down the route of tasteless ed/bookmark/bait maxxing on failure. don't.
making models learn • eXperiments lab • memes and training lores
The sheer magnitude of ambition 🐳: LLMs can prove nontrivial theorems. Let's use PutnamBench. Our best system does 7.4%. Much work remains. everyone: 😮 ooooO it's TIME💯 to hill-climb! Our novel idea of "decomposing problems"… 15%! 35%! 70%! - 🐋: Enough. 118/120, saturated
We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1
Perhaps notably, at the moment of this tweet, the season 1 winner’s losses so far in season 1.5 have more than offset its final gains from season 1
$16,827.71 well spent