MAAT: Multi-phase Adapter-Aware Targeted Unlearning 事件

MAAT: Multi-phase Adapter-Aware Targeted Unlearning arXiv:2605.30514v1 Announce Type: cross Abstract: Machine unlearning evaluation is structurally skewed: Why-type questions, which probe causal and relational knowledge, comprise less than 0.06% of CounterFact, 0.6% of ZSRE, and less than 1.3% of TOFU, MUSE, and WMDP-Cyber. This near-zero representation means that methods that fail on causal knowledge can score highly in aggregate, and this failure is undetectable without balanced evaluation. W