Philippe Chlenski (@pchlenski) 's Twitter Profile
Philippe Chlenski

@pchlenski

cs phd student @ columbia

ID: 986628294102650885

linkhttp://chlenski.com calendar_today18-04-2018 15:31:30

20 Tweet

65 Followers

301 Following

ColumbiaCompSci (@columbiacompsci) 's Twitter Profile Photo

Congratulations to our students and alums who were named #NSFGRFP fellows: Philippe Chlenski (Philippe Chlenski) Udai Nagpal  Shashaank Narayanan Meghna Pancholi  Clayton Sanford (@clayton_sanford) Joseph Zuckerman (Joseph Zuckerman) Alexander Bienstock

bioRxiv (@biorxivpreprint) 's Twitter Profile Photo

MiSDEED: a synthetic multi-omics engine for microbiome power analysis and study design biorxiv.org/cgi/content/sh… #bioRxiv

Bioinformatics Advances (@bioinfoadv) 's Twitter Profile Photo

MiSDEED: a synthetic data engine for microbiome study power analysis and study design by Philippe Chlenski, Melody Hsu, Itsik Pe'er. Bioinformatics Advances, doi.org/10.1093/bioadv…

MiSDEED: a synthetic data engine for microbiome study power analysis and study design by <a href="/pchlenski/">Philippe Chlenski</a>, Melody Hsu, <a href="/Itsikp/">Itsik Pe'er</a>. Bioinformatics Advances, doi.org/10.1093/bioadv…
Jacob Dunefsky (@jacobdunefsky) 's Twitter Profile Photo

Excited to share what I've been working on alongside Philippe Chlenski under Neel Nanda's mentorship: finding interpretable feature circuits in LLMs using a tool called transcoders! Blogpost: lesswrong.com/posts/YmkjnWtZ… Code: github.com/jacobdunefsky/… (1/8)

Jacob Dunefsky (@jacobdunefsky) 's Twitter Profile Photo

Announcing an ICML MI Workshop Spotlight w/ Philippe Chlenski and Neel Nanda! MLP GELUs make it hard to analyze circuits in LLMs. We address this with transcoders, an SAE variant that replaces MLPs with wide, sparse, interpretable layers. This lets us understand complex model behavior!