PhD student @CogSciUCSD | prev RA + visiting student @mitbrainandcog | BITS Pilani '24
Joined June 2023
- Tweets52
- Following231
- Followers85
- Likes2.6K
Chaitanya Kapoor retweeted
Simple input-output dependencies explain neuronal activity (
@ChrisWLynn
) has made the rounds on X with divided opinions. I show in this blog that while neurons are described simply, so can explicitly complex units: Describe and Arise are distinct claims. open.substack.com/pub/anshks…
Super excited to share our latest work Partial Soft-Matching Distance for Neural Representational Comparison with Partial Unit Correspondence in ICLR 2026 with the inimitable @itsneuronal and my phenomenal advisor @meenakshik93!
(1/n)
However, the key difference here is that we know a priori which pair of neural systems are more identical than the other. As seen in the figure, standard soft-matching fails. It prefers model B because noise neurons inflate model A’s score.
(16/n)
However, partial soft-matching recovers the more “similar” model of the two, since there are a larger number of shared signal neurons.
(17/17)
Paper: arxiv.org/abs/2602.19331v1
Code: github.com/NeuroML-Lab/parti…
I’ll be presenting my poster (#2008) from 4:30-7:30p this Friday (5th dec)! Come check it out :)
Excited for @NeurIPSConf 2025! If you’re interested in comparative analysis of neural representations (new tools, why representations align, and large-scale audits of convergence), I’d love to chat. Our lab has several posters this year by @ckapoor07, @andre_longon, Sudhanshu Srivastava, Jialin Wu, @shreya_saha_25, Yiqing Bo, @Ansh_soni1234 @unireps
