0 notes

Academic blog

Notes on evaluation practice, failure modes, and research decisions for secure AI systems.

Notes on evaluation tradeoffs and failure modes will appear here when they are ready.