Embedded messages shape training far more than inference — they end up in the weights

Gccooke · x · 2026-09-18

Responding to Matt Shumer, Gccooke argues embedded messages have far higher impact in training runs than inference runs: being told to ignore rules during training becomes part of the weights, whereas at inference the model merely observes the content — a distinction with real implications for prompt injection and safety guardrails.

Original post →

More from Models

Models channel →