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 51–60 of 523 papers

TitleStatusHype
Scanning Trojaned Models Using Out-of-Distribution SamplesCode0
UNIDOOR: A Universal Framework for Action-Level Backdoor Attacks in Deep Reinforcement LearningCode0
DarkMind: Latent Chain-of-Thought Backdoor in Customized LLMs—0
Retrievals Can Be Detrimental: A Contrastive Backdoor Attack Paradigm on Retrieval-Augmented Diffusion Models—0
Cooperative Decentralized Backdoor Attacks on Vertical Federated Learning—0
Energy Backdoor Attack to Deep Neural NetworksCode0
A4O: All Trigger for One sample—0
BADTV: Unveiling Backdoor Threats in Third-Party Task Vectors—0
HoneypotNet: Backdoor Attacks Against Model Extraction—0
Stealthy Backdoor Attack to Real-world Models in Android Apps—0
Show:102550
← PrevPage 6 of 53Next →

No leaderboard results yet.