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 1–50 of 523 papers

TitleStatusHype
AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge BasesCode3
Agent Security Bench (ASB): Formalizing and Benchmarking Attacks and Defenses in LLM-based AgentsCode3
BAPLe: Backdoor Attacks on Medical Foundational Models using Prompt LearningCode2
BadChain: Backdoor Chain-of-Thought Prompting for Large Language ModelsCode2
An LLM-Assisted Easy-to-Trigger Backdoor Attack on Code Completion Models: Injecting Disguised Vulnerabilities against Strong DetectionCode2
Backdoor Learning: A SurveyCode2
Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based AgentsCode2
Test-Time Backdoor Attacks on Multimodal Large Language ModelsCode2
Few-Shot Backdoor Attacks on Visual Object TrackingCode1
Fast-FedUL: A Training-Free Federated Unlearning with Provable Skew ResilienceCode1
FedDefender: Backdoor Attack Defense in Federated LearningCode1
BadCM: Invisible Backdoor Attack Against Cross-Modal LearningCode1
BadMerging: Backdoor Attacks Against Model MergingCode1
BadHash: Invisible Backdoor Attacks against Deep Hashing with Clean LabelCode1
FIBA: Frequency-Injection based Backdoor Attack in Medical Image AnalysisCode1
Deep Feature Space Trojan Attack of Neural Networks by Controlled DetoxificationCode1
Backdoor Attacks on Self-Supervised LearningCode1
Composite Backdoor Attacks Against Large Language ModelsCode1
Defending Against Backdoor Attacks in Natural Language GenerationCode1
Clean-Label Backdoor Attacks on Video Recognition ModelsCode1
CorruptEncoder: Data Poisoning based Backdoor Attacks to Contrastive LearningCode1
DBA: Distributed Backdoor Attacks against Federated LearningCode1
Embedding and Extraction of Knowledge in Tree Ensemble ClassifiersCode1
Exploring Backdoor Vulnerabilities of Chat ModelsCode1
To Think or Not to Think: Exploring the Unthinking Vulnerability in Large Reasoning ModelsCode1
BadCLIP: Dual-Embedding Guided Backdoor Attack on Multimodal Contrastive LearningCode1
BadEdit: Backdooring large language models by model editingCode1
A new Backdoor Attack in CNNs by training set corruption without label poisoningCode1
BadEncoder: Backdoor Attacks to Pre-trained Encoders in Self-Supervised LearningCode1
Beyond Traditional Threats: A Persistent Backdoor Attack on Federated LearningCode1
CL-Attack: Textual Backdoor Attacks via Cross-Lingual TriggersCode1
Anti-Distillation Backdoor Attacks: Backdoors Can Really Survive in Knowledge DistillationCode1
Anti-Backdoor Learning: Training Clean Models on Poisoned DataCode1
BAGM: A Backdoor Attack for Manipulating Text-to-Image Generative ModelsCode1
BEAGLE: Forensics of Deep Learning Backdoor Attack for Better DefenseCode1
Be Careful about Poisoned Word Embeddings: Exploring the Vulnerability of the Embedding Layers in NLP ModelsCode1
Bkd-FedGNN: A Benchmark for Classification Backdoor Attacks on Federated Graph Neural NetworkCode1
Backdoor Attack against Speaker VerificationCode1
Backdoor Attacks on Crowd CountingCode1
Backdoor Attacks for Remote Sensing Data with Wavelet TransformCode1
Backdoor Attacks on Federated Learning with Lottery Ticket HypothesisCode1
Backdoor Attack on Hash-based Image Retrieval via Clean-label Data PoisoningCode1
Backdoor Attacks Against Dataset DistillationCode1
Backdoor Attacks to Graph Neural NetworksCode1
Can We Mitigate Backdoor Attack Using Adversarial Detection Methods?Code1
Backdoor Attack with Sparse and Invisible TriggerCode1
BadPrompt: Backdoor Attacks on Continuous PromptsCode1
An Embarrassingly Simple Backdoor Attack on Self-supervised LearningCode1
Backdoor Defense via Deconfounded Representation LearningCode1
CleanCLIP: Mitigating Data Poisoning Attacks in Multimodal Contrastive LearningCode1
Show:102550
← PrevPage 1 of 11Next →

No leaderboard results yet.