mlsanityData to 7 Oct 2026

CACTUS: Mask-Guided Semantic Clean-Label Backdoors in Decentralized Federated Learning

arXiv 2609.02450 ↗First posted 2 Sep 2026cs.LG cs.DCcited by 0

Not on the map. Why?

Abstract (arXiv)

Semantic triggers in federated learning (FL) can be less conspicuous than synthetic patches, but sample-dependent placement may weaken backdoor implantation across aggregation rounds. This challenge is compounded in decentralized FL (DFL), where topology-dependent peer aggregation repeatedly mixes local models. CACTUS converts label-consistent semantic pairs into target-directed representation shifts. Mask-guided, modality-specific operators isolate trigger effects, couple them across samples, and apply the shifts counterfactually to clean non-target embeddings before peer aggregation. Experiments cover speech, text, tabular, and image tasks under nine aggregation rules. With 30% malicious nodes, CACTUS reaches a nine-rule mean attack success rate (ASR) of 51.2% on Speech Commands and the highest nine-rule mean ASR among evaluated attacks on three of four modalities. Sensitivity analyses show that ASR varies with network topology and increases with the malicious-node ratio. These results indicate that CACTUS can propagate backdoors through repeated DFL aggregation.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 13 of its 18 references that are in the map. First posted on arXiv 2 Sep 2026.

Builds on

13 of 18 references.

References · 13

  1. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,592
  2. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  3. 2017BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chaincited by 1,052
  4. 2018How To Backdoor Federated Learningcited by 745
  5. 2018Speech Commands: A Dataset for Limited-Vocabulary Speech Recognitioncited by 741
  6. 2018Byzantine-Robust Distributed Learning: Towards Optimal Statistical Ratescited by 698
  7. 2020FLTrust: Byzantine-robust Federated Learning via Trust Bootstrappingcited by 330
  8. 2022Neurotoxin: Durable Backdoors in Federated Learningcited by 97
  9. 2023Backdoor Federated Learning by Poisoning Backdoor-Critical Layerscited by 16
  10. 2024DART: A Solution for Decentralized Federated Learning Model Robustness Analysiscited by 5
  11. 2023Sentinel: An Aggregation Function to Secure Decentralized Federated Learningcited by 4
  12. 2026Beyond Corner Patches: Semantics-Aware Backdoor Attack in Federated Learningcited by 2
1 more
  1. 2025A Crowdsensing Intrusion Detection Dataset For Decentralized Federated Learning Modelscited by 1

Built on by

Not cited yet.