5 Matching Annotations
  1. Jun 2026
    1. These vulnerabilities all appear relatively simple, and we have found that other publicly-available models are able to discover them as well without requiring a bypass.

      Anthropic声称发现的漏洞相对简单且其他模型也存在类似问题,这可能带有淡化政府担忧的偏见。需要独立验证这些漏洞的严重性以及与其他模型的比较是否准确。

  2. Apr 2021
  3. Oct 2020
  4. Jul 2020
  5. May 2020