Event ·Multi-source

GPT-6 Astra在Terminal-Bench-Science基准测试中表现优异

GPT-6 Astra在Terminal-Bench-Science 0.1基准测试中得分为64.6%。这一成绩比GPT-5.6的22.4%提升了42.2个百分点。Terminal-Bench-Science基准测试发布仅一周就被OpenAI采用。该基准测试旨在解决世界 toughest的科学挑战。

Current read

OpenAI刚发布的GPT-6 Astra在科学基准测试中得分暴涨42%，远超前代模型。

11 sources73° AI heatLast updated 2026/9/3 21:56:12

Evidence

11 sources

Primary sourceStanford AI Lab

2026/9/3 21:56:12View source

Related source 1OpenAI

2026/9/3 19:32:16View source

Related source 2elvis

2026/9/3 19:55:42View source

Related source 3Greg Brockman

2026/9/3 20:20:14View source

Related source 4orange.ai

2026/9/3 23:36:56View source

Related source 5IT之家

2026/9/1 22:28:51View source

Related source 6shao__meng

2026/9/1 23:39:07View source

Related source 7Simon Willison’s Weblog

2026/9/1 23:57:28View source

Related source 8The Rundown AI

2026/9/3 18:07:36View source

Related source 9Decoder

2026/9/3 19:25:40View source

Related source 10@OpenAIDevs

2026/9/3 19:33:20View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story