Self-interpretability: LLMs can describe complex internal processes that drive their decisions, and improve with training

Dillon Plunkett, Adam Morris, Keerthi Reddy, Jorge Morales · 2025 · arXiv

Read the paper

Shows that models can describe features of the internal processes behind their own decisions, and that training improves this ability.

Questions it bears on

• How well can LLMs introspect