SOTAVerified

Model extraction

Model extraction attacks, aka model stealing attacks, are used to extract the parameters from the target model. Ideally, the adversary will be able to steal and replicate a model that will have a very similar performance to the target model.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 11–20 of 176 papers

TitleStatusHype
ATOM: A Framework of Detecting Query-Based Model Extraction Attacks for Graph Neural NetworksCode1
Now You See Me (CME): Concept-based Model ExtractionCode1
"Yes, My LoRD." Guiding Language Model Extraction with Locality Reinforced DistillationCode1
Neural Honeytrace: A Robust Plug-and-Play Watermarking Framework against Model Extraction AttacksCode1
Black-Box Attacks on Sequential Recommenders via Data-Free Model ExtractionCode1
Cryptanalytic Extraction of Neural Network ModelsCode1
Entangled Watermarks as a Defense against Model ExtractionCode1
FedRolex: Model-Heterogeneous Federated Learning with Rolling Sub-Model ExtractionCode1
An anatomy-based V1 model: Extraction of Low-level Features, Reduction of distortion and a V1-inspired SOM—0
Adversarial Exploitation of Policy Imitation—0
Show:102550
← PrevPage 2 of 18Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1three-step-originalExact Match0.17—Unverified