SPAR Seeks Mentees for AI Safety Research: Focusing on Metagaming and Eval Awareness

austinc3301 · x · 2026-07-31

Researcher Ezra J. Newman is recruiting mentees on the SPAR platform for an AI safety research project.

The project focuses on 'metagaming' (where models act strategically during evaluations) and 'eval awareness' in advanced AI models. Exceptional fellows will have the opportunity to work at the well-known AI safety organization Apollo Research.

Original post →

More from Safety

Safety channel →