AI 导读
Epoch AI 宣布 FrontierMath Tier 4 基准已饱和。该基准 2025 年 7 月 11 日发布时最高分为 5%，不到 14 个月后升至 98%。
正文
When we launched Tier 4 on July 11th, 2025, the top score was 5%. Less than 14 months later, the top score is 98% and we consider the benchmark saturated. https://t.co/Ta0Cxx5BBn
来源：@EpochAIResearch · x.com