Tag: AI evaluation
Browse our exclusive articles!
HealthBench: AI Evaluation Benchmark for Healthcare Safety
Discover HealthBench, the new AI evaluation benchmark ensuring safety, accuracy, and reliability in healthcare AI models with physician collaboration.
OpenAI & Anthropic Reveal Key AI Safety Evaluation Results
OpenAI and Anthropic share insights from a joint AI safety evaluation focusing on misalignment, hallucinations, and security vulnerabilities.
Detecting and Preventing Scheming in AI Models
Learn how Apollo Research and OpenAI detect and reduce scheming in AI models to improve alignment and ensure safer AI behavior.
Evaluating AI Models with GDPval on Real-World Tasks
Discover how GDPval measures AI performance on 44 real-world, economically valuable tasks to boost productivity and industry impact.
IndQA: Benchmark for AI in 12 Indian Languages
Discover IndQA, OpenAI's new benchmark evaluating AI across 12 Indian languages with cultural and reasoning insights.