今天新增gpt-6VIDEO
RoboColosseum 大规模评测：GPT-6 Astra 在真实场景表现突出
41 天、124 个网站、11398 次真实评测对比 5 个 VLA 策略与 3 个前沿 agent，结论是 GPT-6 Astra 表现很强。
Jiafei Duan✈️CoRL2026
@DJiafei1 小时前收录原推发布于 2026.11.05
Jiafei Duan✈️CoRL2026 @DJiafei原推发布于 2026.11.05
41 days. 124 diverse sites. 11,398 real-world evaluations. At @RoboColosseum, we compared 5 VLA policies and 3 frontier AI agents in the wild. One clear takeaway: @OpenAI’s GPT-6 Astra is seriously strong. Those demos on X line up with what we’re seeing at scale. Read our 34 pages technical report for many more interesting insights: https://arxiv.org/pdf/2611.02497
x.com/djiafei/status/2118149011316326720采集于 2026.11.05 10:05