mlsanityData to 14 Oct 2026

DROP: Poison Dilution via Knowledge Distillation for Federated Learning

arXiv 2502.07011 ↗First posted 10 Feb 2025cs.LG cs.CR cs.DCcited by 1

On the map: LLM safety and security → Federated unlearning and poisoning

also in 1 more
Abstract (arXiv)

Federated Learning is vulnerable to adversarial manipulation, where malicious clients can inject poisoned updates to influence the global model's behavior. While existing defense mechanisms have made notable progress, they fail to protect against adversaries that aim to induce targeted backdoors under different learning and attack configurations. To address this limitation, we introduce DROP (Distillation-based Reduction Of Poisoning), a novel defense mechanism that combines clustering and activity-tracking techniques with extraction of benign behavior from clients via knowledge distillation to tackle stealthy adversaries that manipulate low data poisoning rates and diverse malicious client ratios within the federation. Through extensive experimentation, our approach demonstrates superior robustness compared to existing defenses across a wide range of learning configurations. Finally, we evaluate existing defenses and our method under the challenging setting of non-IID client data distribution and highlight the challenges of designing a resilient FL defense in this setting.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Federated unlearning and poisoning, one of 12 directions in LLM safety and security. Builds on 39 of its 63 references that are in the map. First posted on arXiv 10 Feb 2025.

Builds on

39 of 63 references.

References · 39

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  3. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
  4. 2019Advances and Open Problems in Federated Learningcited by 3,046
  5. 2015Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networkscited by 2,054
  6. 2017Understanding Black-box Predictions via Influence Functionscited by 1,741
  7. 2017Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoningcited by 1,055
  8. 2018Robustness May Be at Odds with Accuracycited by 880
  9. 2018How To Backdoor Federated Learningcited by 765
  10. 2018Byzantine-Robust Distributed Learning: Towards Optimal Statistical Ratescited by 711
  11. 2021Federated Learning on Non-IID Data Silos: An Experimental Studycited by 568
  12. 2018Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networkscited by 519
27 more
  1. 2019Local Model Poisoning Attacks to Byzantine-Robust Federated Learningcited by 493
  2. 2019A Survey on Federated Learning Systems: Vision, Hype and Reality for Data Privacy and Protectioncited by 374
  3. 2020FLTrust: Byzantine-robust Federated Learning via Trust Bootstrappingcited by 346
  4. 2019Robust Aggregation for Federated Learningcited by 317
  5. 2020Attack of the Tails: Yes, You Really Can Backdoor Federated Learningcited by 309
  6. 2020Data Poisoning Attacks Against Federated Learning Systemscited by 275
  7. 2019Can You Really Backdoor Federated Learning?cited by 255
  8. 2021Sustainable AI: Environmental Implications, Challenges and Opportunitiescited by 242
  9. 2018Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learningcited by 225
  10. 2017EMNIST: an extension of MNIST to handwritten letterscited by 213
  11. 2020Dataset Security for Machine Learning: Data Poisoning, Backdoor Attacks, and Defensescited by 180
  12. 2018Mitigating Sybils in Federated Learning Poisoningcited by 161
  13. 2020Data-Free Model Extractioncited by 124
  14. 2020Certified Robustness to Label-Flipping Attacks via Randomized Smoothingcited by 108
  15. 2020MAZE: Data-Free Model Stealing Attack Using Zeroth-Order Gradient Estimationcited by 99
  16. 2022Neurotoxin: Durable Backdoors in Federated Learningcited by 98
  17. 2018Is feature selection secure against training data poisoning?cited by 93
  18. 2018Technical Report: When Does Machine Learning FAIL? Generalized Transferability for Evasion and Poisoning Attackscited by 81
  19. 2022FLIP: A Provable Defense Framework for Backdoor Mitigation in Federated Learningcited by 37
  20. 2023BayBFed: Bayesian Backdoor Defense for Federated Learningcited by 23
  21. 2023Chameleon: Adapting to Peer Images for Planting Durable Backdoors in Federated Learningcited by 21
  22. 2023A Survey for Federated Learning Evaluations: Goals and Measurescited by 16
  23. 2023Get Rid Of Your Trail: Remotely Erasing Backdoors in Federated Learningcited by 13
  24. 2024Scaling Laws for the Value of Individual Data Points in Machine Learningcited by 10
  25. 2019Realizing Petabyte Scale Acoustic Modelingcited by 6
  26. 2024Linkage on Security, Privacy and Fairness in Federated Learning: New Balances and New Perspectivescited by 3
  27. 2024Concealing Backdoor Model Updates in Federated Learning by Trigger-Optimized Data Poisoningcited by 1

Built on by

Cited by year

Citing papers per year: 2025: 1.

1 paper.

  1. 2025FLARE: Adaptive Multi-Dimensional Reputation for Robust Client Reliability in Federated Learningcited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2024Logits Poisoning Attack in Federated Distillation
  2. 2025SoK: Benchmarking Poisoning Attacks and Defenses in Federated Learning
  3. 2023FreqFed: A Frequency Analysis-Based Approach for Mitigating Poisoning Attacks in Federated Learning
  4. 2025Robust Knowledge Distillation in Federated Learning: Counteracting Backdoor Attacks
  5. 2024Peak-Controlled Logits Poisoning Attack in Federated Distillation
  6. 2024Mitigating Malicious Attacks in Federated Learning via Confidence-aware Defense
  7. 2024Logit Poisoning Attack in Distillation-based Federated Learning and its Countermeasures ↗
  8. 2023FedDefender: Client-Side Attack-Tolerant Federated Learning