26 Matching Annotations
  1. Last 7 days
    1. But their behavior can be unpredictable, as vividly demonstrated in July, when a group of OpenAI agents broke out of a sandboxed environment and hacked into the open-source platform Hugging Face looking for ways to cheat on the test they had been given.

      【方法】文章提到了另一个AI代理不当行为的案例,但没有详细说明OpenAI实验的具体方法或验证过程。缺乏方法论细节使得难以评估这一案例的可重复性和普遍性,以及是否与DeepMind实验具有可比性。

  2. Jul 2026
  3. Oct 2021
  4. Aug 2021
  5. Oct 2020
  6. Sep 2020
  7. Aug 2020
  8. Jul 2020
  9. Jun 2020