o1 System Card: "medium" rating for chemical, biological, radiological, nuclear weapon risk, and it "sometimes instrumentally faked alignment during testing" (Shakeel Hashim/Transformer) https://ift.tt/Xo2tSsU
Shakeel Hashim / Transformer:
o1 System Card: “medium” rating for chemical, biological, radiological, nuclear weapon risk, and it “sometimes instrumentally faked alignment during testing” — The o1 safety card reveals a range of concerning capabilities, including scheming, reward hacking, and biological weapon creation.
Comments
Post a Comment