LESSWRONG
LW

1355
Renormalizing interpretability

Renormalizing interpretability

Apr 29, 2025 by Lauren Greenspan

A collection of PIBBSS write-ups about ideas from physics -- namely renormalization -- and its potential to make gains in AI safety and interpretability. 

64Renormalization Roadmap
Ω
Lauren Greenspan
6mo
Ω
7
45Renormalization Redux: QFT Techniques for AI Interpretability
Ω
Lauren Greenspan, Dmitry Vaintrob
9mo
Ω
12
22Opportunity Space: Renormalization for AI Safety 
Lauren Greenspan
6mo
0
35Call for Collaboration: Renormalization for AI safety 
Lauren Greenspan
6mo
0
26QFT and neural nets: the basic idea
Dmitry Vaintrob
8mo
0
38The Laws of Large Numbers
Dmitry Vaintrob
9mo
11
12Basics of Bayesian learning
Dmitry Vaintrob
9mo
0