Kenny Peng
@kennylpeng
CS PhD student at Cornell Tech. Interested in interactions between algorithms and society. Princeton math '22.
The mainstream view of AI for science says AI will rapidly accelerate science, and that we're on track to cure cancer, double the human lifespan, colonize space, and achieve a century of progress in the next decade. In a new AI Snake Oil essay, @random_walker and I argue that…
I've resolved this positively: 2 papers convincingly show sparse autoencoders beating baselines on real tasks: Hypothesis Generation & Auditing LLMs SAEs shine when you don't know what you're looking for, but lack precision. Sometimes the right tool for the job, sometimes not.
Manifold Market: Will Sparse Autoencoders be successfully used on a downstream task in the next year and beat baselines? Stephen Grugett asked me for alignment-relevant markets, this was my best idea. I think SAEs are promising, but how far can they go? manifold.markets/NeelNanda/will…
1. We will present HypotheSAEs at #ICML2025, Wednesday 11am (West Hall B2-B3 #W-421). 2. Let me know if you'd like to chat about: - AI for hypothesis generation - why SAEs are still useful - whether PhD students should stay in school
We'll present HypotheSAEs at ICML this summer! 🎉 We're continuing to cook up new updates for our Python package: github.com/rmovva/Hypothe… (Recently, "Matryoshka SAEs", which help extract coarse and granular concepts simultaneously without as much hyperparameter fiddling.)
💡New preprint & Python package: We use sparse autoencoders to generate hypotheses from large text datasets. Our method, HypotheSAEs, produces interpretable text features that predict a target variable, e.g. features in news headlines that predict engagement. 🧵1/
I’m really excited to share the first paper of my PhD, “Learning Disease Progression Models That Capture Health Disparities” (accepted at #CHIL2025)! Link and summary in thread✨1/