"Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training."

Evan Hubinger et al. (2024)

Details and statistics

DOI: 10.48550/ARXIV.2401.05566

access: open

type: Informal or Other Publication

metadata version: 2024-01-25

a service of  Schloss Dagstuhl - Leibniz Center for Informatics