evald.ai · METR Blog Frontier AI Safety Policies February 08, 2025 16:00 Model Evaluation & Threat Research Open original
