Amin Karbasi
@aminkarbasi
Followers
10K
Following
7K
Media
206
Statuses
3K
Senior director of Cisco Foundation AI, Former Chief Scientist at Robust Intelligence. ex Professor at Yale University, ex staff research scientist at Google.
Joined September 2013
My brilliant student @AnayMehrotra and amazing post-doc Alkis Kalavasis just won the Best Paper Award at COLT 2025! Clearly, they’ve cracked the code: do great math, write a killer paper, take home the trophy. Huge congratulations to both of them. Now please don’t forget your.
13
13
175
I think he means mathematics :).
Physics sees through all lies perfectly.
0
0
9
Unfortunately, the adding information part will lead to so much hallucination that the model may not be able to recover from. @elonmusk, correctly, always talks about following the first principles, asking whether there is a fundamental reason we may be wrong. I think in this.
We will use Grok 3.5 (maybe we should call it 4), which has advanced reasoning, to rewrite the entire corpus of human knowledge, adding missing information and deleting errors. Then retrain on that. Far too much garbage in any foundation model trained on uncorrected data.
1
0
10
“High Dimensional Probability” is one of favorite books of all time. I even taught it once. Now, the second edition is out.
2
36
271
Let machines become you!.
Don’t become a machine.
1
0
3
RT @FannyYangETH: Register now (first-come first-served) for the "Math of Trustworthy ML workshop" at #LagoMaggiore, Switzerland, Oct 12-16….
0
21
0
Mark Twain said "history rhymes". It is not that it repeats itself, it rhymes. I would say the same about AI benchmark crisis.
0
0
1
An amazing list of speakers.
Invited talked are announced. Jon Kleinberg.Pamela Samuelson.Frauke Kreuter.Anca Dragan.Andreas Krause.
0
0
10
I also have AI winter by the way
0
0
6
AI summer
0
0
10
Together with @HamedSHassani , @AlexRobey23 , and @yaronsinger , we will be presenting an @icmlconf 2025 tutorial:. “Jailbreaking LLMs and Agentic Systems: Attacks, Defenses, and Evaluations.”. Key takeaways:.--Latest jailbreak techniques—from prompt injections to agent exploits.
1
5
45
“I learned very early the difference between knowing the name of something and knowing something.” Richard Feynman.
2
0
24
RT @AlexKontorovich: At IAS lunch yesterday:. "It's OK that AI will soon be better at math than humans, we'll survive -- just look at chess….
0
19
0
RT @kasperglarsen: I have an opening for a post doc position in machine learning theory, with a deadline of June 1st. Please share and appl….
0
9
0
We just released: Foundation-Sec-8B, an open-weight, 8-billion-parameter base model purpose-built for cybersecurity. Why it’s special:.State-of-the-art in a small package: performs on par with—or even ahead of—models many times its size (e.g., Llama-3.1-70B) on CTI-MCQA and.
1
5
32
There is a certain beauty in walking alone.
1
1
23