Model Confidence Under Answer-Preserving Attacks: An Informativeness-Manipulability Frontier

Add code
Aug 06, 2026

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: