causal
In-context learning as System 2-like pattern extraction
In-context learning, where a model infers a task from a few examples such as the Marie Curie completion, is a general routine that arises from prediction training, works after weights are frozen, and resembles the rapid, novel pattern-grokking attributed to System 2.
causalpending
Speaker
Chandra SripadaEvidence Quote
“It's a surprising phenomenon that prediction training installs in the model. Something like a very general routine for extracting patterns from the prompt”
Created: 11/17/2026, 6:53:48 AM
Your notes
Loading notes...