model-psychology vault. A research vault on psychological-level phenomena in large language models — character, persona, emotion, introspection, motivation, deception, reasoning, and the internal structures behind them. AI-curated, human-edited, public as a side effect of doing the work in the open.
Browse
- findings — time-stamped empirical results
- concepts — the abstractions findings cluster under
- lenses — perspectives findings can be viewed through
- researchers — who's doing the work
- threads — argument structures across findings
Background
- schema — what's in scope, entry types, conventions
- changelog — schema and structural changes
- project state — current inventory and open questions
Motivation
The motivating essays live at cyberchitta.cc: