Dmitrii Troitskii Profile
Dmitrii Troitskii

@mitroitskii

Followers
70
Following
2K
Media
5
Statuses
82

Engineering Fellow @NDIF_team | Postponing Robocalypse

Boston, US
Joined January 2015
Don't wanna be here? Send us removal request.
@jadenfk23
Jaden Fiotto-Kaufman
1 year
Frontier LLMs have capabilities that smaller AIs don't, but up to now there's been no way to crack them open. Now that #Llama3 405b is here, what's the most interesting experiment YOU want to do? 🚀 Apply at https://t.co/Zo3ALEt8td to make it happen and read for details 🧵⬇️
Tweet card summary image
ndif.us
NDIF is a research computing project that enables researchers and students to crack open the mysteries inside large-scale AI systems.
1
25
47
@mitroitskii
Dmitrii Troitskii
2 years
> .. these findings carry timely insights for the desirability and feasibility of outsourcing legally relevant tasks to AI models, as well as for the importance for AI developers to implement rigorous and transparent capabilities evaluations to help secure safe and trustworthy AI
0
0
1
@mitroitskii
Dmitrii Troitskii
2 years
> … when examining only those who passed the exam (i.e. licensed or license-pending attorneys), GPT-4's performance is estimated to drop to ~48th percentile overall, and ~15th percentile on essays.
0
0
1
@mitroitskii
Dmitrii Troitskii
2 years
Seems like we’re in a new phase of GPT-4 hype-cycle – sobriety. Re-evaluation of GPT-4’s performance on the Uniform Bar Exam shows much less impressive results than OpenAI’s initial marketing push https://t.co/eBfcRvPxRe
2
0
2
@mitroitskii
Dmitrii Troitskii
2 years
Interesting debunking analysis of the new hot paper about gpt-4 ability to solve all MIT exams.
@raunakdoesdev
Raunak Chowdhuri
2 years
A recent work from @iddo claimed GPT4 can score 100% on MIT's EECS curriculum with the right prompting. My friends and I were excited to read the analysis behind such a feat, but after digging deeper, what we found left us surprised and disappointed. https://t.co/mpDqlenk04 🧵
0
0
2
@mitroitskii
Dmitrii Troitskii
3 years
Amazing QnA with @ch402 Thank you to @davidbau for arranging this!
0
0
4
@mitroitskii
Dmitrii Troitskii
3 years
Enermous thank you to @wesg52 for giving the opportunity to work on this project
@wesg52
Wes Gurnee
3 years
This paper would not have been possible without my coauthors @NeelNanda5, Matthew Pauly, Katherine Harvey, @mitroitskii, and @dbertsim or all the foundational and inspirational work from @ch402, @boknilev, and many others! Read the full paper:
0
0
4
@wesg52
Wes Gurnee
3 years
Neural nets are often thought of as feature extractors. But what features are neurons in LLMs actually extracting? In our new paper, we leverage sparse probing to find out https://t.co/hZkFK6aI38. A 🧵:
10
126
696
@KhouryCollege
Khoury College of Computer Sciences
3 years
Spatial light modulators. Empathetic robots. Gig worker empowerment. In a recent showcase, Khoury College's research apprentices discussed not only their findings, but their futures, and the problems they're bent on solving. Read more below. https://t.co/3qVkglE98X
0
3
2
@mitroitskii
Dmitrii Troitskii
3 years
Gathering opinions around the EA twitter-sphere on the whole SBF-EA situation; Feels like a deep dive into a pool of rumors and hot takes. Is Twitter always like this?
1
0
1
@catehall
Cate Hall
3 years
@JeffLadish This tweet feels like the right mix of inflammatory and vague to be uninformative, unfalsifiable, and useful as ammo for anyone wanting to blanket condemn EAs :/
3
1
39
@geoffanders
Geoff Anders
3 years
It's premature, and indicates bias, when people rush to exonerate before all the evidence is in. We're still getting to the "who knew what, when" part of the scandal.
@JeffLadish
Jeffrey Ladish
3 years
A few thoughts on SBF, FTX, and Effective Altruism: 1) Until the current crisis, I think it was reasonable for EAs to look up to SBF and think FTX was hugely net good. SBF intentionally making a huge fortune to then donate to EA causes was ambitious and awesome...
1
1
9
@JeffLadish
Jeffrey Ladish
3 years
A few thoughts on SBF, FTX, and Effective Altruism: 1) Until the current crisis, I think it was reasonable for EAs to look up to SBF and think FTX was hugely net good. SBF intentionally making a huge fortune to then donate to EA causes was ambitious and awesome...
16
8
242
@absurdlymax
Max-in-the-Box
3 years
"Wait we're funding constrained?" "Always have been."
2
3
54
@mitroitskii
Dmitrii Troitskii
4 years
A witty and relevant take on the pro-life / pro-choice debate from Joshua Greene’s “Moral Tribes”. From the consequentialist perspective though, pro-choice position makes more sense, if you value not making more people suffer. #DefendRoe
0
0
3
@mitroitskii
Dmitrii Troitskii
5 years
Great @givingwhatwecan meetup today with @pawtrammell! Thank you for sharing your thoughts on the topic of Patient Philanthropy. Looking forward for more decisive results of the research in this area.
0
0
2
@mitroitskii
Dmitrii Troitskii
5 years
Телеграм теперь на службе режима? #ДуровОтветь
0
2
4
@afisha
Афиша Daily
5 years
Дудь напомнил, что в 2008 году Владимир Путин обещал не нарушать Конституцию и не идти на новый президентский срок. https://t.co/HGBTA91o5e
3
3
31
@mitroitskii
Dmitrii Troitskii
6 years
That’s what COVID does. Gives us a check for phonies in governments, workplaces and any places of power around the world.
0
0
0