返回首页
RadarAI··论文与技术

Artificial Analysis 推出 HLE-Diamond 榜单,Gemini 4 Carbon (high) 以 69% 居首

中文摘要

Artificial Analysis 发布 HLE-Diamond 榜单,评测精炼版 Humanity's Last Exam,Gemini 4 Carbon (high) 以 69% 夺冠。

English Summary

Artificial Analysis launched the HLE-Diamond leaderboard evaluating refined Humanity's Last Exam questions, with Gemini 4 Carbon (high) leading at 69%.

原文节选

Artificial Analysis 推出 HLE-Diamond 榜单,对 CAIS 和 Scale AI 清理精炼后的 1000 题 Humanity's Last Exam 版本做独立评测,纳入 Artificial Analysis Intelligence Index v5。