MMODELYST
People/Robin Jia
PPL

Robin Jia

Papers (3)
Spiking the training data to correct for test set contaminationSHRED: Retain-Set-Free Unlearning via Self-Distillation with Logit DemotionHow Language Models Process Negation