Xuewei Yang

Delta-MIA: Measuring Membership Inference Attacks in Large Language Models via self-Contrast Framework

S2J: Bridging the Gap Between Solving and Judging Ability in Generative Reward Models