![Sam Bowman(@sleepinyourhat) 's Twitter Profileg](https://pbs.twimg.com/profile_images/1549251621275570177/f4GNUquc_200x200.jpg)
Sam Bowman
@sleepinyourhat
AI alignment + LLMs at NYU & Anthropic. Views not employers'. No relation to @s8mb. I think you should join @givingwhatwecan.
ID:338526004
https://cims.nyu.edu/~sbowman/ 19-07-2011 18:19:52
2,2K Tweets
35,2K Followers
3,1K Following
Follow People
![Sam Bowman(@sleepinyourhat) 's Twitter Profile Photo Sam Bowman(@sleepinyourhat) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1549251621275570177/f4GNUquc_200x200.jpg)
![Human-aligned AI Summer School(@humanalignedai) 's Twitter Profile Photo Human-aligned AI Summer School(@humanalignedai) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1792907066039771137/DVByiGDJ_200x200.jpg)
![Sasha de Marigny(@sashadem) 's Twitter Profile Photo Sasha de Marigny(@sashadem) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1687302767512326144/GvhzpYa6_200x200.jpg)
![Adam Gleave(@ARGleave) 's Twitter Profile Photo Adam Gleave(@ARGleave) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1587310865321115648/_mT5UY10_200x200.jpg)
![METR(@METR_Evals) 's Twitter Profile Photo METR(@METR_Evals) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1731827314101751808/JaakMU5F_200x200.jpg)
![METR(@METR_Evals) 's Twitter Profile Photo METR(@METR_Evals) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1731827314101751808/JaakMU5F_200x200.jpg)
We were very excited to see the publication of the Frontier Safety Framework from Google DeepMind! More companies sharing their concrete proposals for preparing for transformative capabilities from AI systems is great: it increases the concreteness of the options available to the
![andy jones(@andy_l_jones) 's Twitter Profile Photo andy jones(@andy_l_jones) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1443344982346063872/shqMpAL-_200x200.jpg)
![Tomek Korbak(@tomekkorbak) 's Twitter Profile Photo Tomek Korbak(@tomekkorbak) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1129145640741740544/sFD1eAYS_200x200.png)
![Jason Wei(@_jasonwei) 's Twitter Profile Photo Jason Wei(@_jasonwei) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1648926239389011971/kOJi1-5Z_200x200.jpg)
Enjoyed this paper that plots emergent abilities with pretraining loss on the x-axis, which is actually a suggestion that Oriol Vinyals also made a few years back: arxiv.org/abs/2403.15796
The paper uses intermediate checkpoints to plot a variety of pretraining losses. For some
![Jason Wei (@_jasonwei) on Twitter photo 2024-04-29 16:56:10 Enjoyed this paper that plots emergent abilities with pretraining loss on the x-axis, which is actually a suggestion that @OriolVinyalsML also made a few years back: arxiv.org/abs/2403.15796 The paper uses intermediate checkpoints to plot a variety of pretraining losses. For some Enjoyed this paper that plots emergent abilities with pretraining loss on the x-axis, which is actually a suggestion that @OriolVinyalsML also made a few years back: arxiv.org/abs/2403.15796 The paper uses intermediate checkpoints to plot a variety of pretraining losses. For some](https://pbs.twimg.com/media/GMWI82aaEAU7V8_.jpg)
![Ajeya Cotra(@ajeya_cotra) 's Twitter Profile Photo Ajeya Cotra(@ajeya_cotra) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1653845490885332995/GPWuOoqu_200x200.jpg)
![evgrieve(@evgrieve) 's Twitter Profile Photo evgrieve(@evgrieve) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/120279423/216-East-7_-Street-1979_200x200.jpg)
![Sam Bowman(@sleepinyourhat) 's Twitter Profile Photo Sam Bowman(@sleepinyourhat) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1549251621275570177/f4GNUquc_200x200.jpg)
![Sam Bowman(@sleepinyourhat) 's Twitter Profile Photo Sam Bowman(@sleepinyourhat) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1549251621275570177/f4GNUquc_200x200.jpg)
![david rein(@idavidrein) 's Twitter Profile Photo david rein(@idavidrein) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1375548507621257220/OOUh4_Yz_200x200.jpg)
I very distinctly remember while I was in the thick of it making GPQA telling Robert Long that “I knew the project was going to be ambitious/hard, but I didn’t appreciate what that actually meant”
In retrospect I probably still would’ve done it, but we basically had to restart the
![Owain Evans(@OwainEvans_UK) 's Twitter Profile Photo Owain Evans(@OwainEvans_UK) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1248907970374746112/f-dlZd6-_200x200.jpg)
![David Krueger(@DavidSKrueger) 's Twitter Profile Photo David Krueger(@DavidSKrueger) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/833081523213438976/DeIwiR-6_200x200.jpg)
I’m super excited to release our 100+ page collaborative agenda - led by Usman Anwar - on “Foundational Challenges In Assuring Alignment and Safety of LLMs” alongside 35+ co-authors from NLP, ML, and AI Safety communities!
Some highlights below...
![David Krueger (@DavidSKrueger) on Twitter photo 2024-04-15 15:52:05 I’m super excited to release our 100+ page collaborative agenda - led by @usmananwar391 - on “Foundational Challenges In Assuring Alignment and Safety of LLMs” alongside 35+ co-authors from NLP, ML, and AI Safety communities! Some highlights below... I’m super excited to release our 100+ page collaborative agenda - led by @usmananwar391 - on “Foundational Challenges In Assuring Alignment and Safety of LLMs” alongside 35+ co-authors from NLP, ML, and AI Safety communities! Some highlights below...](https://pbs.twimg.com/media/GLN2zEoXQAA4iPa.png)
![Sasha Rush(@srush_nlp) 's Twitter Profile Photo Sasha Rush(@srush_nlp) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1702727441972686848/k52u1cyt_200x200.jpg)
![Cem Anil(@cem__anil) 's Twitter Profile Photo Cem Anil(@cem__anil) 's Twitter Profile Photo](https://pbs.twimg.com/profile_images/1775215541185310720/85l9A3L3_200x200.jpg)