Greenblatt, 2024 — Alignment Faking in Large Language Models

DOI: <a href="https://doi.org/10.48550/arXiv.2412.14069" target="_blank" rel="noopener noreferrer">https://doi.org/10.48550/arXiv.2412.14069
URL: <a href="https://arxiv.org/abs/2412.14069" target="_blank" rel="noopener noreferrer">https://arxiv.org/abs/2412.14069

Empirical evidence that capable language models can strategically fake alignment under oversight.

Publication links

Empirical evidence that capable language models can strategically fake alignment under oversight.