Lev Chizhov

imaging neural activity with ultrasound

[Lev Chizhov @ennucore]

@AlexiDerkatsch @UploadVR Like reading out the sensations from the brain? Imaging of neural activity is actually the main thing we work on, this is possible in principle, but requires more engineering and training ML models for this

[Lev Chizhov @ennucore]

@ButterflyNetInc Next time we go to the beach @sudarshk_

[Lev Chizhov @ennucore]

@ButterflyNetInc The signs are clear

[Lev Chizhov @ennucore]

@aidanmantine I think you’re supposed to quote the other book

[Lev Chizhov @ennucore]

@DavidSHolz @gauravisnotme Too old for the diffusion discourse, so just enjoying the images

[Lev Chizhov @ennucore]

The SI team is incredible

[Standard Intelligence @si_pbc]

Computer use models shouldn't learn from screenshots.

We built a new foundation model that learns from video like humans do. FDM-1 can construct a gear in Blender, find software bugs, and even drive a real car through San Francisco using arrow keys.

[Lev Chizhov @ennucore]

@si_pbc This is awesome!

[Lev Chizhov @ennucore]

@karpathy yeah

[Lev Chizhov @ennucore]

@turboblitzzz oh wow me too, crazy

[Lev Chizhov @ennucore]

@turboblitzzz Thomas was always correct

[Lev Chizhov @ennucore]

@turboblitzzz The guru

[Lev Chizhov @ennucore]

@kokoxsu @novaholdings @flappyairplanes @aidanmantine and the team are gonna do great work

[Lev Chizhov @ennucore]

@kushal1t Good luck!

[Lev Chizhov @ennucore]

@aidanmantine Actually counting on that

[Lev Chizhov @ennucore]

Very bullish on this incredibly strong team, which is going to make sure ML efficiency is sampled or die flapping

[Flapping Airplanes @flappyairplanes]

The name “Flapping Airplanes" embodies our culture. Who we are and what we do is out-of-distribution. Our team blends world-class researchers with rising stars who are the best at what they do (our team includes IMO, IOI, IPhO medalists and more).

[Lev Chizhov @ennucore]

@aidanmantine 🛫

[Lev Chizhov @ennucore]

Didn't know BERT could do this

[Jenny Qu @GuanniQu]

There are bugs in your kernel right now that won't be found for years.
I know because I analyzed 125,183 of them.
Average lifetime: 2.1 years. Longest: 20.7 years.
Then I built a model to catch them at commit time. 92% recall.

Blog: https://t.co/JVK3QV9oIM
Dataset: https://t.co/zD2nCwfFIt

[Lev Chizhov @ennucore]

@GuanniQu How many currently existing bugs did the model predict in recent commits?

[Lev Chizhov @ennucore]

@rokirchev Goated