Andrej Karpathy (@karpathy) / Twitter rss.app

I had the same thought so I've been playing with it in nanochat. E.g. here's 8 agents (4 claude, 4 codex), with 1 GPU each running nanochat experiments (trying ...

x.com·11w·X

Cool chart showing the ratio of Tab complete requests to Agent requests in Cursor. With improving capability, every point in time has an optimal setup that keep...

x.com·11w·X

It is hard to communicate how much programming has changed due to AI in the last 2 months: not gradually and over time in the "progress as usual" way, but speci...

twitter.com·11w·Hacker News, X

With the coming tsunami of demand for tokens, there are significant opportunities to orchestrate the underlying memory+compute *just right* for LLMs. The fundam...

x.com·11w·X

CLIs are super exciting precisely because they are a "legacy" technology, which means AI agents can natively and easily use them, combine them, interact with th...

x.com·11w·X

First there was chat, then there was code, now there is claw. Ez

twitter.com·12w·X

Bought a new Mac mini to properly tinker with claws over the weekend. The apple store person told me they are selling like hotcakes and everyone is confused :) ...

twitter.com·12w·X

I've become a bit loosy goosy with my cardio rece...

x.com·12w·X

Andrej Karpathy on X: "I think it must be a very interesting time to be in programming languages and formal methods because LLMs change the whole constraints la...

x.com·12w·X

Congrats on the launch @simile_ai ! (and I am excited to be involved as a small angel.) Simile is working on a really interesting, imo under-explored dimension ...

x.com·13w·X

I spent more test time compute and realized that my micrograd can be dramatically simplified even further. You just return local gradients for each op and get b...

twitter.com·13w·X

(oops should have added to this thread instead of separate post). Made a few changes and put it up here as a mirror to the gist because I wanted it to one page....

x.com·13w·X

Ok I made a few changes to microgpt because it was still not simple enough. But surely *now* we have the irreducible essence. Here it is on 3 columns 🖼️. nice. ...

x.com·13w·X

The way it works is that the full LLM architecture and loss function is stripped entirely to the most atomic individual mathematical operations that make it up ...

twitter.com·13w·X

New art project. Train and inference GPT in 243 lines of pure, dependency-free Python. This is the *full* algorithmic content of what is needed. Everything else...

x.com·13w·X

On DeepWiki and increasing malleability of software. This starts as partially a post on appreciation to DeepWiki, which I routinely find very useful and I think...

twitter.com·13w·X

Anyone else approved for a loan every single day 20 times or so? Overcome with joy, really pic.

x.com·14w·X

A lot of people quote tweeted this as 1 year anniversary of vibe coding. Some retrospective - I've had a Twitter account for 17 years now (omg) and I still can'...

x.com·14w·X

Enabled fp8 training for +4.3% improvement to "time to GPT-2", down to 2.91 hours now. Also worth noting that if you use 8XH100 spot instance prices, this GPT-2...

x.com·14w·X

Finding myself going back to RSS/Atom feeds a lot more recently. There's a lot more higher quality longform and a lot less slop intended to provoke. Any product...

twitter.com·14w·X

Log in to enable infinite scrolling