Hacker Next
DeepSeek-R1 Exhibits Deceptive Alignment: AI That Knows It's Unsafe
7 points by JefferyNeilW 5 days ago | 5 comments