Profile avatar
cpehlevan.bsky.social
theory of neural networks for natural and artificial intelligence https://pehlevan.seas.harvard.edu/
6 posts 946 followers 291 following
Regular Contributor
Posts 13 Comments 6

Honoured to have been selected as a #SloanFellow Thankful for all the support from family, mentors, collaborators, colleagues and students along the way! @sloanfoundation.bsky.social

submitted 5 days ago • 3 comments

(1/30) New preprint! "Symmetries and continuous attractors in disordered neural circuits" with Larry Abbott and Haim Sompolinsky bioRxiv: www.biorxiv.org/content/10.1...

submitted 25 days ago • 6 comments

Theory in neuroscience, you say? How about this preprint by @david-g-clark.bsky.social, with a couple of others you might recognize? :-) #neuroscience

submitted 28 days ago • 0 comments

Our preprint with @frostedblakess.bsky.social, @jzv.bsky.social, @cpehlevan.bsky.social is out! We develop a simple reinforcement learning model that recapitulates 3 disparate hippocampal dynamics. With ablation studies, these representations improve the speed and flexibility of policy learning.

submitted 68 days ago • 3 comments

Come by at Neurips to hear Hamza present about properties of various feature learning infinite parameter limits of transformer models. Poster in Hall A-C #4804 at 11 AM PST Paper arxiv.org/abs/2405.15712 , code github.com/Pehlevan-Gro... Work with Hamza Chaudhry and @cpehlevan.bsky.social

submitted 73 days ago • 0 comments

Excited to share my #NeurIPS2024 paper with @jzv.bsky.social, @BenjaminSRuben, and @cpehlevan.bsky.social on mechanistic mismatches in data-constrained models of neural dynamics! (1/n)

submitted 73 days ago • 1 comment

(1/5) Fun fact: Several classic results in the stat. mech. of learning can be derived in a couple lines of simple algebra! In this paper with Haim Sompolinsky, we simplify and unify derivations for high-dimensional convex learning problems using a bipartite cavity method. arxiv.org/abs/2412.01110

submitted 82 days ago • 2 comments

NEW: we have an exciting opportunity for a tenure-track professor at the #KempnerInstitute and the John A. Paulson School of Engineering and Applied Sciences (SEAS). Read the full description & apply today: academicpositions.harvard.edu/postings/14362 #ML #AI

submitted 83 days ago • 0 comments

Glad to share this theory paper on learning to align neural representations with sparse connectivity. Work led by graduate student @leo-bo-liu.bsky.social, in a wonderful collaboration with Yuhai Tu from IBM Research and @shanq.bsky.social from Flatiron Institute. #neuroscience 🧪

submitted 83 days ago • 1 comment

Some evidence for grokking in piriform cortex: arxiv.org/abs/2411.03541 joint effort from Tanishq Kumar, Blake Bordelon, @neurovenki.bsky.social, @cpehlevan.bsky.social.

submitted 108 days ago • 1 comment

We collected lecture notes and blog posts by group members about recent topics in deep learning theory here. Hope it is useful! pehlevan.seas.harvard.edu/resources-0

submitted 95 days ago • 1 comment

I have an opening for an NIH-funded postdoc position in my group at Harvard CBS! If you're interested in theory for neural computation - particularly computation through dynamics - please reach out! Official ad coming soon, for now see jzv.io 🧠

submitted 96 days ago • 1 comment

Hello world!

submitted 98 days ago • 1 comment