OpenAI将定期公布其模型安全评估结果

Core Viewpoint - OpenAI has launched a safety assessment center for its models, aiming to enhance safety transparency and provide periodic updates on its security efforts [1] Group 1 - The safety assessment center will regularly update its content as part of OpenAI's initiative to strengthen safety transparency [1] - The results from the center will reflect only certain dimensions of OpenAI's safety work, providing a snapshot rather than a comprehensive evaluation [1] - A complete assessment of model safety and performance should combine data from this center with system cards, framework evaluation reports, and specialized research reports released with each model [1]