Evaluation and Testing¶
Methods for evaluating AI systems in safety-critical contexts — benchmarking, prototyping, HRA adaptation, and assessment frameworks.
- hra-methods-for-ai — Adapting human reliability analysis methods to AI-inclusive sociotechnical systems
- ai-specific-performance-shaping-factors — Factors that shape human performance when working with AI systems
- capability-gradient — Assessing AI capability across task types, complexity levels, and operational conditions