Explore by topic
Papers as research projects, and notes on the questions behind them.
Papers & Projects
Research stories, methods, and results behind my papers and ongoing projects.
Cognition
Concepts, learning, reasoning, and the structure of minds.
AI
Language models, mechanistic understanding, and intelligent agents.
Research
Scientific questions, methodology, and the practice of doing research.
Latest notes
-
A Clean Neuron is Not Yet an Explanation
Why progress in reading individual neurons leaves the central questions of mechanistic interpretability open.
-
Using Mechanistic Interpretability for Cognitive Science
What can looking inside artificial neural networks teach us about computation, intelligence, and human cognition?