Are LLMs Vulnerable to Preference-Undermining Attacks (PUA)? A Factorial Analysis Methodology for Diagnosing the Trade-off between Preference Alignment and Real-World Validity
Paper • 2601.06596 • Published • 12 •
4
liangliang
Fi-Liang
·
AI & ML interests
None yet
Organizations
None yet
commented a
paper 6 months ago