Noam Brown· @polynoamial · X·· 2026-05-12AI 评分61
AI 导读
趣闻：这些致命错误最初是用@OpenAI的GPT-5.5标记的 [引用 @EpochAIResearch]：我们正在对FrontierMath的1-4级进行AI辅助审查。这已标记出约三分之一题目的致命错误，且我们认为大多数标记是有效的。完成人工审核后，我们将在修正数据集上公布更新分数。
正文 · 原文
Fun fact: the fatal errors were initially flagged using @OpenAI's GPT-5.5
We are conducting an AI-assisted review of FrontierMath: Tiers 1-4. This has flagged fatal errors in about a third of problems, and we believe most of these flags to be valid. We will release updated scores on a corrected dataset after completing a thorough human review.在 X 查看被引用的帖子
来源：Noam Brown · x.com