backdoor attacks - Robuta Search

https://openreview.net/forum?id=cemEOP8YoC IBA: Towards Irreversible Backdoor Attacks in Federated Learning | OpenReview Federated learning (FL) is a distributed learning approach that enables machine learning models to be trained on decentralized data without compromising end... backdoor attacks federated learning iba towards irreversible https://openreview.net/forum?id=t0ozPUGnBs UMD: Unsupervised Model Detection for X2X Backdoor Attacks | OpenReview Backdoor (Trojan) attack is a common threat to deep neural networks, where samples from one or more source classes embedded with a backdoor trigger will be... backdoor attacks umd unsupervised model detection https://openreview.net/forum?id=VNMJfBBUd5&referrer=%5Bthe%20profile%20of%20Baoyuan%20Wu%5D(%2Fprofile%3Fid%3D~Baoyuan_Wu1) Activation Gradient based Poisoned Sample Detection Against Backdoor Attacks | OpenReview This work studies the task of poisoned sample detection for defending against data poisoning based backdoor attacks. Its core challenge is finding a... backdoor attacks activation gradient based poisoned https://openreview.net/forum?id=QmrHMnzJG0- Backdoor Attacks Against Transformers with Attention Enhancement | OpenReview We propose a novel Trojan Attention Loss (TAL), which enhances the Trojan behavior by directly manipulating the attention pattern. backdoor attacks transformers attention enhancement openreview https://deepai.org/publication/rethink-stealthy-backdoor-attacks-in-natural-language-processing Rethink Stealthy Backdoor Attacks in Natural Language Processing | DeepAI Jan 9, 2022 - 01/09/22 - Recently, it has been shown that natural language processing (NLP) models are vulnerable to a kind of security threat called the B... natural language processing backdoor attacks rethink stealthy deepai https://deepai.org/publication/deepsight-mitigating-backdoor-attacks-in-federated-learning-through-deep-model-inspection DeepSight: Mitigating Backdoor Attacks in Federated Learning Through Deep Model Inspection | DeepAI Jan 3, 2022 - 01/03/22 - Federated Learning (FL) allows multiple clients to collaboratively train a Neural Network (NN) model on their private data without... backdoor attacks federated learning https://openreview.net/forum?id=i-k6J4VkCDq Marksman Backdoor: Backdoor Attacks with Arbitrary Target Class | OpenReview Backdoor Attacks with Arbitrary Target Class backdoor attacks marksman arbitrary target class https://deepai.org/publication/detecting-backdoor-attacks-on-deep-neural-networks-by-activation-clustering Detecting Backdoor Attacks on Deep Neural Networks by Activation Clustering | DeepAI Nov 9, 2018 - 11/09/18 - While machine learning (ML) models are being increasingly trusted to make decisions in different and varying areas, the safety of ... deep neural networks backdoor attacks detecting https://openreview.net/forum?id=3vr3TXMRFo&referrer=%5Bthe%20profile%20of%20Yuan%20Yao%5D(%2Fprofile%3Fid%3D~Yuan_Yao7) Inspecting Prediction Confidence for Detecting Black-Box Backdoor Attacks | OpenReview Backdoor attacks have been shown to be a serious security threat against deep learning models, and various defenses have been proposed to detect whether a... black box backdoor attacks inspecting prediction confidence https://openreview.net/forum?id=BfVkbfJGW4 Graph Contrastive Backdoor Attacks | OpenReview Graph Contrastive Learning (GCL) has attracted considerable interest due to its impressive node representation learning capability. Despite the wide... backdoor attacks graph contrastive openreview https://www.sify.com/tag/backdoor-attacks/ backdoor attacks Archives - Sify backdoor attacks archives sify https://aclanthology.org/2023.findings-emnlp.716/ Attention-Enhancing Backdoor Attacks Against BERT-based Models - ACL Anthology Weimin Lyu, Songzhu Zheng, Lu Pang, Haibin Ling, Chao Chen. Findings of the Association for Computational Linguistics: EMNLP 2023. 2023. backdoor attacks attention enhancing bert based https://deepai.org/publication/hidden-killer-invisible-textual-backdoor-attacks-with-syntactic-trigger Hidden Killer: Invisible Textual Backdoor Attacks with Syntactic Trigger | DeepAI May 26, 2021 - 05/26/21 - Backdoor attacks are a kind of insidious security threat against machine learning models. After being injected with a backdoor in ... backdoor attacks hidden killer invisible textual https://www.techradar.com/pro/security/thousands-of-wordpress-sites-targeted-with-malicious-plugin-backdoor-attacks Thousands of WordPress sites targeted with malicious plugin backdoor attacks | TechRadar Mar 7, 2025 - JavaScript was deploying four separate backdoors on vulnerable websites wordpress sites backdoor attacks thousands targeted https://openreview.net/forum?id=rkgyS0VFvr DBA: Distributed Backdoor Attacks against Federated Learning | OpenReview We proposed a novel distributed backdoor attack on federated learning and show that it is not only more effective compared with standard centralized attacks,... backdoor attacks federated learning dba distributed openreview https://thehackernews.com/2023/02/north-korean-cyber-espionage-group.html Cyber Espionage Group Earth Kitsune Deploys WhiskerSpy Backdoor in Latest Attacks North Korean group hacked a pro-North Korean organization's website and distributed their latest backdoor, WhiskerSpy, to unsuspecting visitors. cyber espionage group earth kitsune https://arxiv.org/abs/2411.12701 [2411.12701] When Backdoors Speak: Understanding LLM Backdoor Attacks Through Model-Generated... Abstract page for arXiv paper 2411.12701: When Backdoors Speak: Understanding LLM Backdoor Attacks Through Model-Generated Explanations https://openreview.net/forum?id=bntZBG9fBY Multimodal Deception in Explainable AI: Concept-Level Backdoor Attacks on Concept Bottleneck Models... Deep learning has demonstrated transformative potential across domains, yet its inherent opacity has driven the development of Explainable Artificial... explainable ai https://openreview.net/forum?id=wZLWuFHxt5 A Survey of Recent Backdoor Attacks and Defenses in Large Language Models | OpenReview Large Language Models (LLMs), which bridge the gap between human language understanding and complex problem-solving, achieve state-of-the-art performance on... https://fossbytes.com/baidus-corrupt-sdk-puts-100-million-android-users-open-to-backdoor-attacks/ Baidu's Corrupt SDK Puts 100 Million Android Users Open To Backdoor Attacks Nov 3, 2015 - Baidu, the Chinese developer has distributed a software development kit Moplus, which has put backdoors to 100 million Android users exposing https://thehackernews.com/2024/04/russian-apt-deploys-new-kapeka-backdoor.html Russian APT Deploys New 'Kapeka' Backdoor in Eastern European Attacks A new stealthy backdoor malware called Kapeka, likely created by Russia's APT group Sandworm, has been targeting Eastern Europe. eastern european russian apt deploys new https://thehackernews.com/2024/10/north-korean-hackers-using-new.html North Korean Hackers Using New VeilShell Backdoor in Stealthy Cyber Attacks North Korean hackers target Southeast Asia with VeilShell malware, a sophisticated backdoor used for covert operations. north korean hackers using new https://www.tripwire.com/state-of-security/patch-your-sony-ip-cameras-against-backdoor-attacks Patch Your Sony Security Cameras Against Backdoor Attacks! If you have a Sony network-connected CCTV camera, you may have a security problem. security cameras patch sony backdoor attacks https://www.tripwire.com/state-of-security/bad-actors-using-mitm-attacks-against-asus-to-distribute-plead-backdoor Bad Actors Using MitM Attacks against ASUS to Distribute Plead Backdoor Researchers believe bad actors are using man-in-the-middle (MitM) attacks against ASUS software to distribute the Plead backdoor. bad actors mitm attacks using https://arxiv.org/abs/2502.20650 [2502.20650] Gungnir: Exploiting Stylistic Features in Images for Backdoor Attacks on Diffusion... Abstract page for arXiv paper 2502.20650: Gungnir: Exploiting Stylistic Features in Images for Backdoor Attacks on Diffusion Models https://thehackernews.com/2022/06/chinese-luoyu-hackers-using-man-on-side.html Chinese LuoYu Hackers Using Man-on-the-Side Attacks to Deploy WinDealer Backdoor LuoYu, an "extremely sophisticated" Chinese-speaking APT hacker group, has been observed using man-on-the-side attacks to deploy a Windows backdoor. man on the side https://openreview.net/forum?id=jIAPLDdGVx Provable Robustness of (Graph) Neural Networks Against Data Poisoning and Backdoor Attacks |... Generalization of machine learning models can be severely compromised by data poisoning, where adversarial changes are applied to the training data. This... graph neural networks https://www.helpnetsecurity.com/2015/12/21/week-review-screenos-backdoor-twitter-warns-state-sponsored-attacks-mackeeper-data-breach/ Week in review: ScreenOS backdoor, Twitter warns about state-sponsored attacks, and MacKeeper data... Oct 15, 2018 - Here's an overview of some of last week's most interesting news and articles: Critical ScreenOS bugs allow undetectable decryption of VPN connections, week in review https://thehackernews.com/2022/09/ta505-hackers-using-teslagun-panel-to.html TA505 Hackers Using TeslaGun Panel to Manage ServHelper Backdoor Attacks Researchers detail TeslaGun, a previously undocumented software control panel used by the financially motivated cybercrime group TA505. to manage ta505 hackers using panel https://deepai.org/publication/dangerous-cloaking-natural-trigger-based-backdoor-attacks-on-object-detectors-in-the-physical-world Dangerous Cloaking: Natural Trigger based Backdoor Attacks on Object Detectors in the Physical... Jan 21, 2022 - 01/21/22 - Deep learning models have been shown to be vulnerable to recent backdoor attacks. A backdoored model behaves normally for inputs c... https://arxiv.org/abs/2301.13862 [2301.13862] Salient Conditional Diffusion for Defending Against Backdoor Attacks Abstract page for arXiv paper 2301.13862: Salient Conditional Diffusion for Defending Against Backdoor Attacks 2301 13862 salient conditional diffusion https://www.umassd.edu/events/cms/5-25-26-a-noise-based-defense-for-stealthy-backdoor-attacks.php Events: A Noise-Based Defense for Stealthy Backdoor Attacks in Large Vision-Language Models | UMass... May 25, 2026 to May 25, 2026