SOTAVerified

Machine Unlearning

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–25 of 438 papers

TitleStatusHype
What Should LLMs Forget? Quantifying Personal Data in LLMs for Right-to-Be-Forgotten Requests—0
Model State Arithmetic for Machine UnlearningCode0
On the Necessity of Output Distribution Reweighting for Effective Class Unlearning—0
Recalling The Forgotten Class Memberships: Unlearned Models Can Be Noisy Labelers to Leak Privacy—0
Verifiable Unlearning on Edge—0
Orthogonal Soft Pruning for Efficient Class Unlearning—0
Sharpness-Aware Machine Unlearning—0
Unlearning Isn't Invisible: Detecting Unlearning Traces in LLMs from Model OutputsCode1
Membership Inference Attacks as Privacy Tools: Reliability, Disparity and EnsembleCode1
Rectifying Privacy and Efficacy Measurements in Machine Unlearning: A New Inference Attack PerspectiveCode1
UCD: Unlearning in LLMs via Contrastive Decoding—0
Apollo: A Posteriori Label-Only Membership Inference Attack Towards Machine UnlearningCode0
Prompt Attacks Reveal Superficial Knowledge Removal in Unlearning Methods—0
SoK: Machine Unlearning for Large Language Models—0
Certified Unlearning for Neural NetworksCode1
Towards Lifecycle Unlearning Commitment Management: Measuring Sample-level Unlearning CompletenessCode0
Quantifying Cross-Modality Memorization in Vision-Language Models—0
Rethinking Machine Unlearning in Image Generation ModelsCode1
Invariance Makes LLM Unlearning Resilient Even to Unanticipated Downstream Fine-TuningCode0
Towards Machine Unlearning for Paralinguistic Speech Processing—0
SALAD: Systematic Assessment of Machine Unlearing on LLM-Aided Hardware Design—0
Speech Unlearning—0
Aligned but Blind: Alignment Increases Implicit Bias by Reducing Awareness of RaceCode0
Does Machine Unlearning Truly Remove Model Knowledge? A Framework for Auditing Unlearning in LLMs—0
Targeted Unlearning Using Perturbed Sign Gradient Methods With Applications On Medical Images—0
Show:102550
← PrevPage 1 of 18Next →

No leaderboard results yet.