Amin Karbasi
@aminkarbasi
Followers
11K
Following
7K
Media
211
Statuses
3K
Senior director of Cisco Foundation AI, Former Chief Scientist at Robust Intelligence (acquired by Cisco). ex Professor at Yale, ex researcher at Google.
Joined September 2013
I heard plenty of similar stories about him during my time at Yale. At this rate, we might end up replacing Serge Lang with our favorite LLM. The bigger shift, though, is access: what used to be reserved for a small circle of people who knew the “Langs” of the world will soon be
I would call that the "Serge Lang in the next office moment": when I was a postdoc at Yale, my office was next to his. As I was working on my first significant paper, I got stuck on a commutative algebra lemma. After wasting weeks, I knocked on Serge's door and asked him. In 1
0
0
39
I fully agree with this take by @ErnestRyu. When we started our little fun project, we were surprised. Now, we’re seeing more cases of frontier models succeeding on the Gödel Test and solving simple (or maybe not so simple) conjectures. https://t.co/bV3DMN9bIl
arxiv.org
Recent announcements from frontier AI model labs have highlighted strong results on high-school and undergraduate math competitions. Yet it remains unclear whether large language models can solve...
I firmly believe we are at a watershed moment in the history of mathematics. In the coming years, using LLMs for math research will become mainstream, and so will Lean formalization, made easier by LLMs. (1/4)
0
0
10
Evidence is mounting that frontier models can pass the Gödel Test and tackle simple conjectures. Mathematics could be on the cusp of real change.
arxiv.org
Recent announcements from frontier AI model labs have highlighted strong results on high-school and undergraduate math competitions. Yet it remains unclear whether large language models can solve...
I crossed an interesting threshold yesterday, which I think many other mathematicians have been crossing recently as well. In the middle of trying to prove a result, I identified a statement that looked true and that would, if true, be useful to me. 1/3
0
1
6
Tired to go back to the original papers again and again? Our monograph: a systematic and fundamental recipe you can rely on! 📘 We’re excited to release 《The Principles of Diffusion Models》— with @DrYangSong, @gimdong58085414, @mittu1204, and @StefanoErmon. It traces the core
41
414
2K
Readers across America call The Epoch Times “factual, detailed, and unlike anything else.” No spin, no shouting—just clarity. 👇
0
2
12
If you're a PhD student interested in interning with me or one of my amazing colleagues at Microsoft Research New England (@MSRNE, @MSFTResearch) this summer, please apply here https://t.co/DIkXUuK4zc (If you'd like to work with me, please include my name in your cover letter!)
8
71
427
As PhD students, we believe research automation systems should belong to everyone, not just Google, so we built freephdlabor. Customize your multi-agent system for end-to-end research that WORKS FOR YOUR DOMAIN within hours. full source code: https://t.co/NkiFnLwqVM
3
5
7
When we worked on the Cell to Sentence paper two years ago, it was hard to imagine this result. Back then, we were only training tiny models like the ones you fine here https://t.co/hZ2DrXLsDO By scaling the model to 27b parameters and more data, the same method generated a
An exciting milestone for AI in science: Our C2S-Scale 27B foundation model, built with @Yale and based on Gemma, generated a novel hypothesis about cancer cellular behavior, which scientists experimentally validated in living cells. With more preclinical and clinical tests,
0
5
33
In this 6-min video, I show how to use iDEP to interpret bulk RNA-seq data. Start with QC plots and exploratory analyses before identifying differentially regulated genes and pathways. Here, we picked up on high mitochondrial rRNA counts, one male sample mixed in with seven
0
9
49
Youtube is the best university. The best minds upload their best work for free.
0
1
13
Not All Bits Are Equal: What We Learned From 1700 Experiments on Memory-Optimal Reasoning Given a fixed memory budget, how should you allocate across model weights, KV cache, and test-time compute to maximize accuracy in reasoning models? For example: would you choose a 32B,
11
63
435
(1/n) Until recently, even strong LLMs struggled with USAMO/IMO problems. This year, specific model variants from Google and OpenAI’s were reported to solve 5/6 IMO problems. In our recent work, we asked a relevant question: Can we grade proofs fairly with partial credit using
1
21
94
FIRST CLIP from Roses on the Vine. Starring Cameron Cowperthwaite (FALLOUT) and Tonatiuh (KISS OF THE SPIDER WOMAN). Available on UK digital platforms this Monday 11/3. 🌹
1
3
15
Hi new followers! I’m a mathematician at Harvard. I have a YouTube channel, where I discuss math the way I think about it, with now two playlists on differential geometry and complex geometry (in progress). Comments welcome! DG: https://t.co/ywUKaiYfc8 CG: https://t.co/hBwZQFZkHC
19
76
569
No way yo find parking spots on the Berkeley campus anymore.
0
0
8
Soon there will be no parking left for simple professors at Berkeley. No pressure @AlexGDimakis
Congratulations to UC Berkeley’s Omar Yaghi, who shares the 2025 @NobelPrize in #Chemistry for helping create a field called reticular chemistry.
0
0
21
What a journey.
This year’s chemistry laureate Omar Yaghi was born in Amman, Jordan, in 1965 to parents who were refugees from Palestine. When we spoke to him he shared his story: “I grew up in a very humble home, we were a dozen of us in one room, sharing it with the cattle that we used to
0
0
3
Evolutionary Maladaptation & The Promethean Imperative We're running Stone Age software on a civilization that requires quantum computing. Your brain evolved for survival in groups of 20-150 people, handling immediate physical threats and managing scarcity. Now you're processing
10
14
20
Do something worth calling.
0
0
3
The most immediate and reliable value of LLMs in mathematics seems to lie in discovering counterexamples, not in constructing rigorous proofs. This aligns with what both Terrence Tao’s experiments and our own Gödel Test results show. GPT-5 refuted our conjecture, which turned out
4
7
53
A pleasure to be invited by Lisa Blackwell, granddaughter of Prof. David Blackwell, to say a few words about her grandfather at an event this afternoon to promote an upcoming documentary about his life. To quote my colleague Mike Jordan, Blackwell's 'approachability' theory is
4
28
219
Die empty
there are essays, and then there are "violently shift your entire perception of reality" essays
0
2
5
wrote by real humans.
Our book on conditional gradients and Frank-Wolfe methods with Gábor Braun, @ACarderera, @CyrilleCmbt, Hamed Hassani, @aminkarbasi and Aryan Mokhtari just appeared in the MOS-SIAM Series on Optimization #ml #optimization #ai
https://t.co/VMNrLTv8ZO
0
0
4