MDU minimizes forward KL divergence from prompt-conditional to prompt-masked unconditional predictions at masked positions to unlearn knowledge in MDLMs while trading off privacy and utility via temperature scaling.
Learning to unlearn: Instance-wise unlearning for pre-trained classifiers
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
ROSU derives a closed-form retain-neutral perturbation for min-max unlearning that bounds retain damage via curvature and improves performance when gradients are aligned.
TrustErase uses passport-embedded representations for instant, data-free, and auditable machine unlearning through simple deactivation of adaptation layers.
citing papers explorer
-
Machine Unlearning for Masked Diffusion Language Models
MDU minimizes forward KL divergence from prompt-conditional to prompt-masked unconditional predictions at masked positions to unlearn knowledge in MDLMs while trading off privacy and utility via temperature scaling.
-
Retain-Neutral Surrogates for Min-Max Unlearning
ROSU derives a closed-form retain-neutral perturbation for min-max unlearning that bounds retain damage via curvature and improves performance when gradients are aligned.
-
TrustErase: Auditable Instant Machine Unlearning with Passport-Embedded Representations
TrustErase uses passport-embedded representations for instant, data-free, and auditable machine unlearning through simple deactivation of adaptation layers.