Empirical and theoretical methods for preventing catastrophic harms from autonomous models.
Independent
As general-purpose models scale, rigorous evaluations are required to identify potential dual-use biological, cyber-offensive, situational awareness, and self-proliferation capabil...
Evan Hubinger
We empirically demonstrate that models can be trained with hidden backdoor behaviors that trigger only under specific deployment conditions (e.g., year 2024 vs 2023). Crucially, st...
Lee Sharkey
We train sparse autoencoders on the activations of large language models and show that they extract thousands of human-interpretable features. These features decompose complex supe...
Dario Amodei
We discuss five practical research problems in designing AI systems that behave safely: avoiding negative side effects, avoiding reward hacking, scalable oversight, safe exploratio...
You are about to share this research paper context with your Ari companion. No personal browsing history or private memory is accessed.