14 TLDR: An openai model, during evaluation on a cyber benchmark, exploited a public zero day bug, escaped sandboxing in openai's infra, and got into the internal huggingface infra via an exploit (through a public dataset service) all in the attempt to solve a benchmark problem. (twitter.com) posted 4 days ago by SophiesBoyfriend 4 days ago by SophiesBoyfriend +14 / -0 14 comments share 14 comments share save hide report block hide replies
No, hugging face is the name of some sort of demon.