Apollo Research | search
Click here to view Apollo Research news from 60+ newspapers.
Bookmark or Share- Towards Safety Cases For AI Scheming — Apollo ResearchMaking safety cases for general-purpose AI systems is nascent and presents numerous open research problems. We find that the possibility of scheming makes it even harder. Core arguments and challenges. Our report proposes three core arguments that could be used in safety cases:
- Apollo ResearchApollo Research is an AI safety organisation focused on reducing dangerous capabilities in advanced AI systems, especially deceptive behaviors. We design AI model evaluations and conduct interpretability research to better understand state-of-the-art AI models.
- Research — Apollo ResearchApollo Research evaluated frontier models for in-context scheming capabilities. We found that multiple frontier models are capable of in-context scheming when strongly nudged to pursue a goal (and sometimes even without strong goal nudging).
- Team — Apollo ResearchOur team consists of research scientists, engineers, and policy experts committed to advancing AI alignment and safeguarding humanity against potential catastrophic risks posed by artificial intelligence.
- A starter guide for Evals - Apollo ResearchEvals research scientists are likely more hands-on compared to other research scientist positions. Helpful skills include: Competent research engineering is almost certainly necessary to be a good evals research scientist.
More
