Benjamin S Ruben

PhD Candidate in Biophysics @Harvard: Theoretical Neuroscience, Machine Learning Theory, Chromosome Biophysics

[Benjamin S Ruben @BenjaminSRuben]

Check out my latest work, studying a fundamental trade-off in machine learning: should you train a single large model, or an ensemble of smaller models? A huge thanks to my advisor @CPehlevan and collaborators @wl_tong and @hamzatchaudhry.

https://t.co/CIetU02CtS

[Billy Qian @b_qian_]

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

[Stat.ML Papers @StatMLPapers]

Nadaraya-Watson kernel smoothing as a random energy model https://t.co/fNlraiDjP1

[Alex Atanasov @ABAtanasov]

[1/n] Thrilled that this project with @jzavatoneveth and @cpehlevan is finally out! Our group has spent a lot of time studying high dimensional regression and its connections to scaling laws. All our results follow easily from a single central theorem 🧵
https://t.co/A0Dh1iNT4q

[Benjamin S Ruben @BenjaminSRuben]

Come by poster #1004 at the NeurIPS main conference to discuss ensembling methods, noise, and heterogeneity in neural networks!

Video: https://t.co/tdS3qdrVBN

(15/n, fin.)

[Benjamin S Ruben @BenjaminSRuben]

We study the generalization error of ensembled linear regression in the presence of label noise, feature noise, and a “readout noise” which is applied to the predictions of each ensemble member independently. (2/n)

[Benjamin S Ruben @BenjaminSRuben]

I owe huge thank you to my advisor @CPehlevan for help and advice on this project every step of the way. Thank you also to the many members of the Pehlevan group who gave help and advice which made this project possible. (14/n)

[Benjamin S Ruben @BenjaminSRuben]

Can you prevent double-descent without relying on regularization? In a new #NeurIPS2023 paper with my advisor @CPehlevan, we introduce “heterogeneous” ensembles as an efficient method to mitigate double-descent. 🧵(1/n)

ArXiv: https://t.co/guOUMnoALO

[Alex Atanasov @ABAtanasov]

Very happy to share this work in NeurIPS 2023 with @vyasnikhil96, @blake__bordelon, Sab Sainathan, @DepenKenpachi, and @CPehlevan on the consistent behavior of feature-learning networks across large widths https://t.co/XQB6EsAZGO. What is large width consistency? Read on! 1/n