Varun Simha

SWE. X/Twitter activity is my own.

[Varun Simha @BVarunsimha]

@0xmads 👀

[Varun Simha @BVarunsimha]

@zaneczepek 193

[Varun Simha @BVarunsimha]

@pepemoonboy You missed the obvious: “Never bet against Elon!”. Unless you think he is actually the creator!

[Varun Simha @BVarunsimha]

@readswithravi @naval

[Varun Simha @BVarunsimha]

@AMAZlNGNATURE

[Varun Simha @BVarunsimha]

@tekbog Epic!

[Varun Simha @BVarunsimha]

The birth of JARVIS!

[Deedy @deedydas]

I will never read a research paper the same way again.

With Gemini 2.0 Live, I can ask
"What's this paper about?"
"Can you explain this complex equation simply?"
"What does this diagram mean?"

Research is often intentionally abstruse, but now normal people can read them!

[Varun Simha @BVarunsimha]

@RayGonzalezX @MarioNawfal Especially the re-entry and landing !

[Varun Simha @BVarunsimha]

@AMAZlNGNATURE Thanos!

[Varun Simha @BVarunsimha]

@thedankoe 1

[Varun Simha @BVarunsimha]

@0xRaduan @shl Yes, very well!

[Varun Simha @BVarunsimha]

@elonmusk @lexfridman Yessir🙏

[Varun Simha @BVarunsimha]

@jxmnop https://t.co/0DkY15ZVkm

[Varun Simha @BVarunsimha]

https://t.co/0DkY15ZVkm

[Jack Morris @jxmnop]

one of the most important things I know about deep learning I learned from this paper: "Pretraining Without Attention"

this what I found so surprising:
these people developed an architecture very different from Transformers called BiGS, spent months and months optimizing it and training different configurations, only to discover that at the same parameter count, a wildly different architecture produces identical performance to transformers

this may imply that as long as there are enough parameters, and things are reasonably well-conditioned (i.e. a decent number of nonlinearities and and connections between the pieces) then it really doesn't matter how you arrange them, i.e. any sufficiently good architecture works just fine

i feel there's something really deep here, and we may be already very close to the upper bound of how well we can approximate a given function given a certain amount of compute. so we should spend more time thinking about other questions, such as what that function should actually look like (what data? which objective function?) and how to make it more efficient

[tobi lutke @tobi]

This is the oddest package for this excellent message that I can imagine.

[Varun Simha @BVarunsimha]

@space_tintin "space tintin" give your brain space! 😂

[Varun Simha @BVarunsimha]

@jijosunny This is awesome! And I didn't know @arrahman is in tech now! Haha

[Varun Simha @BVarunsimha]

@fchollet Via Negativa!

[Varun Simha @BVarunsimha]

@MasculineTheory *alone.

[Varun Simha @BVarunsimha]

@ShaanVP @TonyRobbins The one and only! 🙏