Measuring the performance of our models on real-world tasks 文章

OpenAI Blog2025-09-25BLOGen

摘要

OpenAI introduces GDPval, a new evaluation that measures model performance on real-world economically valuable tasks across 44 occupations.

相关事件查看全部 (1)

相关人物

暂无数据

相关技术

暂无数据