SOTAVerified

Backdoor Attack

Backdoor attacks inject maliciously constructed data into a training set so that, at test time, the trained model misclassifies inputs patched with a backdoor trigger as an adversarially-desired target class.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 181–190 of 523 papers

TitleStatusHype
Motif-Backdoor: Rethinking the Backdoor Attack on Graph Neural Networks via MotifsCode0
Data Free Backdoor AttacksCode0
Few-shot Backdoor Attacks via Neural Tangent KernelsCode0
AnywhereDoor: Multi-Target Backdoor Attacks on Object DetectionCode0
Adversarial examples are useful too!Code0
DBIA: Data-free Backdoor Injection Attack against Transformer NetworksCode0
How to Craft Backdoors with Unlabeled Data Alone?Code0
Exploiting the Vulnerability of Large Language Models via Defense-Aware Architectural BackdoorCode0
BagFlip: A Certified Defense against Data PoisoningCode0
EmInspector: Combating Backdoor Attacks in Federated Self-Supervised Learning Through Embedding InspectionCode0
Show:102550
← PrevPage 19 of 53Next →

No leaderboard results yet.