Twitxy Saved

Jeff Horwitz

@JeffHorwitz

Reachable via non-Meta owned devices on cell/Signal at (202) 480-4453. Also at jeffhorwitz.com.
Oakland, CA Joined June 2011
2.8K Following 18K Followers 299 Posts

Latest posts 5

Jeff Horwitz @JeffHorwitz
This would be a great relief and I can't tell if it's one very smart guy heralding a breakthrough or one very smart guy's optimistic take that everyone who'd like to be optimistic is desperately clinging to. Time will tell!
Nick Verified account @nickcammarata Oct 15
interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had

with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works

already-clean.goodfire.com
View quoted post on X