noyesno
Joined 13 years ago
Karma
1,296
Submissions
204
Member since
Jun 19
Reach me at elegant.job5108 via fastmail.com
View on Hacker NewsJoined 13 years ago
”1. Chinese LLMs produce more vulnerable code when prompted with a U.S. government persona than without—and the vulnerabilities are highly obfuscated.
2. Chinese LLMs inject PRC-aligned political bias into both the answers and code they generate. ”