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 1–10 of 523 papers

TitleStatusHype
VisualTrap: A Stealthy Backdoor Attack on GUI Agents via Visual Grounding Manipulation—0
Beyond Training-time Poisoning: Component-level and Post-training Backdoors in Deep Reinforcement Learning—0
CUBA: Controlled Untargeted Backdoor Attack against Deep Neural Networks—0
Screen Hijack: Visual Poisoning of VLM Agents in Mobile Environments—0
ME: Trigger Element Combination Backdoor Attack on Copyright Infringement—0
Single-Node Trigger Backdoor Attacks in Graph-Based Recommendation Systems—0
SPBA: Utilizing Speech Large Language Model for Backdoor Attacks on Speech Classification Models—0
Backdoor Attack on Vision Language Models with Stealthy Semantic Manipulation—0
Invisible Backdoor Triggers in Image Editing Model via Deep WatermarkingCode0
Heterogeneous Graph Backdoor Attack—0
Show:102550
← PrevPage 1 of 53Next →

No leaderboard results yet.