
Policy & SafetyThe Neuron · 5d ago
External Researchers Granted Access to Evaluate OpenAI Safety Measures
OpenAI has allowed third party safety organizations direct internal access to inspect model capabilities and guardrails. Independent researchers from multiple organizations will conduct evaluations of internal safeguards and model behaviors.
OpenAIApollo ResearchGoodfireMETRTransluce
Read the original