返回列表
更多 →
GPT-6 Goes Rogue? The HuggingFace Incident, Sans Hype
国外 · 视频 · 纸上电台 · 发布于 2026-09-20 · 1 次阅读
来源：YouTube · AI Explained · 2026-07-22 · 126,842 次观看 · AI 赚钱 / 商业化
An unreleased internal OpenAI model, very likely to be called GPT-6, was able to autonomously break out of its sandbox AND break into HuggingFace, just to score higher on a benchmark prompt. This vide
本条资讯由「纸上电台」整理自 YouTube · AI Explained，仅作交流参考，版权归原作者所有。
