Keep on to blur preview images; turn off to show them clearly
RT @danielkoeth: Blue Skies Forever 💙
Market Design/Entrepreneurship Professor @HarvardHBS & Faculty Affiliate @Harvard Economics; Research @a16zcrypto; Editor @restatjournal; Econ @Quora; … | #QED
What's going on there, how do you screw this up we've seen DS-MoEs scaling gracefully and predictably from 3B to 1T with trivial allometric changes, half of Chinese tech reports is doing just that
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
When you run parallel agents in the same folder without git worktrees
Open-source ecosystem for high-leverage AI builders. GitHub's #1 AI Agents Repo with 72k+ stars. Join 200k+ active AI builders.
RT @elithrar: $5 @Cloudflare Workers plan + $5 @PlanetScale dev node sounding like a winning combination for building the next big thing 😎
vp developers & ai @cloudflare ✨ and how does that error make you feel?
The nature is healing.
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
someone tell me what i'm missing here, because the titled claim seems trivially false to me: they define an LLM as a function that maps sequence s in V^k to vector in R^d assume hidden state in n-bit precision. at some point, there are more inputs possible than hidden states: |V|^k > 2^{n * d} k > n d log(2) / log |V| let's take GPT-2: n=16, d=768, V≈50,000 then collisions *must* happen starting at a context window size of 214 tokens this seems actually kind of bad, right?
phd research @cornell // language models, information theory, science of AI