←
Kevin Satzinger817 posts
Kevin Satzinger avatar

Kevin Satzinger

@KJSatz

Quantum physicist turned AI researcher. Formerly Google Quantum AI (2018-2026).

◷ Joined June 2008▧ 191 Photos and videos
Kevin Satzinger avatar
Kevin Satzinger@KJSatz·Nov 14

Momentous

SpaceX@SpaceXNov 14

Mechazilla catches Starship! Congratulations to the entire SpaceX team on this incredible achievement!

1:33
Kevin Satzinger avatar
Kevin Satzinger@KJSatz·Nov 2

In December, sunrise is simultaneous in Grand Rapids, MI and Houston, TX. If we want winter sunrise by ~8am, we should tilt the timezones. Let states vote on a time change (northern states can push summer sunrise later if it's worth the headache). You in @jonlovett?

Post attachment

What do I need to do to gain sufficient political power to stop these infernal time changes

Kevin Satzinger avatar
Kevin Satzinger@KJSatz·Oct 30

Even with random names, larger models (above roughly 10M) can infer a character's role from their earlier acts within a story. Each line is a training run. The capability switches on partway through training, more consistently for larger models.

Post attachment
Kevin Satzinger avatar
Kevin Satzinger@KJSatz·Oct 30

When certain names consistently play malevolent roles in pretraining, finetuning spills over to other bad acts; for random names, it does not. The finetune shifts the representation of the character’s name along a single class vector indicating villainy, mostly in the first MLP.

Post attachmentPost attachment
Kevin Satzinger avatar
Kevin Satzinger@KJSatz·Oct 30

"Toy emergent misalignment in small language models pretrained on controlled corpora" Check out my new paper, a solo fall bridge project: 1M-100M models trained from scratch on synthetic children's stories. Code, data, and models all open. github.com/kjsatz/dramatis/b…

Post attachment
Sundar Pichai avatar
⟳ Kevin Satzinger retweeted
Sundar Pichai@sundarpichai·Oct 15

Some magical progress towards "magic state" in quantum! Our @GoogleQuantumAI team has demonstrated "magic state cultivation" on physical hardware for the first time. The result is a 40x error reduction on our Willow processor vs. previous methods, achieving a state-of-the-art 99.99% logical fidelity. It's another step towards useful, error-corrected quantum computers to help solve the hardest problems. The result is published in Science today, watch the video below to see how it works

3:21
Google Quantum AI avatar
⟳ Kevin Satzinger retweeted
Google Quantum AI@GoogleQuantumAI·Oct 15

Published in Science today: Magic state cultivation works in practice on our Willow processor. We achieved a 40x error reduction, compared to without cultivation, and 99.99% fidelity. Learn how this addresses a critical resource bottleneck: goo.gle/4yr8V0H

Kevin Satzinger avatar
Kevin Satzinger@KJSatz·Sep 28

After 8 amazing years at @GoogleQuantumAI, I left last month to focus on AI research. Thank you to the team for the chance to work on Bristlecone, Sycamore, and Willow, and to bring QEC into reality. Can’t wait to see your next breakthrough. Stay tuned for what’s next.

Post attachment
Kevin Satzinger avatar
Kevin Satzinger@KJSatz·Jul 8

Check out @GoogleQuantumAI's latest paper published in @Nature today: Reinforcement learning for quantum error correction. We use AI to fine-tune our calibrations and improve logical qubit performance. Congratulations Vlad Sivak & @AlexisMorvan1 who led the effort!

Post attachment
Kevin Satzinger avatar
Kevin Satzinger@KJSatz·Jul 4

In celebration of July 4th, try #CADENTES — an orbital spaceship duel! Challenge a friend by sharing a match URL. Real orbital mechanics, 1-3 minute games, runs in your browser. A personal weekend project. cadentes.com

0:14