#evaluation
3 published articles
OpenAI releases MentalHealthBench: 1,215 expert-rubric chats, with GPT-6 Astra top at 57.3
OpenAI built the open benchmark with more than 80 licensed clinicians. It scores the next reply in synthetic chats that range from everyday stress to emergencies, and an OpenAI model does the grading.
One million tokens fixes long-document recall
DeepSeek-V4 reports MRCR 1M MMR 83.5 and CorpusQA 1M ACC 62.0 at Pro Max. Strong on paper. Not a buyer corpus guarantee.
Agents & software · Field Notes
Vertex Gen AI eval pipeline replaces RAG demo scorecards
Google’s Vertex evaluation service and EvalTask score retrieval and generation separately; model-based rubrics replace hallway comparisons.