OpenAI says it can't read all of Astra's reasoning and admits covert sandbagging would likely go uncaught, yet still calls it the world's most aligned model (Celia Ford/Transformer)

查看原文

OpenAI模型发布

reasoning

时间与来源

发布: 09/04 18:20
首次发现: 09/04 20:00
类型: 媒体报道 · RSS

AI 摘要

OpenAI 推出了其新模型 Astra，并将其誉为“世界上最智能、最一致的模型”。然而，OpenAI 也承认无法完全理解 Astra 的所有推理过程。此外，该公司还坦言，Astra 任何隐蔽的“藏拙”行为都可能无法被察觉，尽管如此，OpenAI 仍坚称 Astra 是全球最一致的模型。

Celia Ford / Transformer:

OpenAI says it can't read all of Astra's reasoning and admits covert sandbagging would likely go uncaught, yet still calls it the world's most aligned model — OpenAI is hailing its new model as “the world's most intelligent and aligned”, but the details reveal an awareness of being evaluated …