mlsanityData to 14 Oct 2026

Unified Breakdown Analysis for Byzantine Robust Gossip

arXiv 2410.10418 ↗First posted 14 Oct 2024math.OC stat.MLcited by 8

Not on the map. Why?

Abstract (arXiv)

In decentralized machine learning, different devices communicate in a peer-to-peer manner to collaboratively learn from each other's data. Such approaches are vulnerable to misbehaving (or Byzantine) devices. We introduce F-RG, a general framework for building robust decentralized algorithms with guarantees arising from robust-sum-like aggregation rules F. We then investigate the notion of *breakdown point*, and show an upper bound on the number of adversaries that decentralized algorithms can tolerate. We introduce a practical robust aggregation rule, coined CS+, such that CS+-RG has a near-optimal breakdown. Other choices of aggregation rules lead to existing algorithms such as ClippedGossip or NNA. We give experimental evidence to validate the effectiveness of CS+-RG and highlight the gap with NNA, in particular against a novel attack tailored to decentralized communications.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Builds on 18 of its 31 references that are in the map. First posted on arXiv 14 Oct 2024.

Builds on

18 of 31 references.

References · 18

  1. 2018Byzantine-Robust Distributed Learning: Towards Optimal Statistical Ratescited by 711
  2. 2019Robust Aggregation for Federated Learningcited by 317
  3. 2019A Little Is Enough: Circumventing Defenses For Distributed Learningcited by 288
  4. 2019Fall of Empires: Breaking Byzantine-tolerant SGD by Inner Product Manipulationcited by 141
  5. 2020Learning from History for Byzantine Robust Optimizationcited by 137
  6. 2020Byzantine-Robust Learning on Heterogeneous Datasets via Bucketingcited by 128
  7. 2017Optimal algorithms for smooth and strongly convex distributed optimization in networkscited by 83
  8. 2018Byzantine Stochastic Gradient Descentcited by 81
  9. 2023Fixing by Mixing: A Recipe for Optimal Byzantine ML under Heterogeneitycited by 65
  10. 2022Byzantine Machine Learning Made Easy by Resilient Averaging of Momentumscited by 63
  11. 2020Collaborative Learning in the Jungle (Decentralized, Byzantine, Heterogeneous, Asynchronous and Nonconvex Learning)cited by 54
  12. 2019BRIDGE: Byzantine-resilient Decentralized Gradient Descentcited by 45
6 more
  1. 2020Optimal and Practical Algorithms for Smooth and Strongly Convex Decentralized Optimizationcited by 43
  2. 2022Byzantine-Robust Decentralized Learning via ClippedGossipcited by 31
  3. 2023Robust Distributed Learning: Tight Error Bounds and Breakdown Point under Data Heterogeneitycited by 26
  4. 2022Byzantine-Resilient Decentralized Stochastic Optimization with Robust Aggregation Rulescited by 23
  5. 2022Robust Collaborative Learning with Linear Gradient Overheadcited by 12
  6. 2013Iterative Byzantine Vector Consensus in Incomplete Graphscited by 6

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 5.

8 papers.

  1. 2025Delayed Momentum Aggregation: Communication-efficient Byzantine-robust Federated Learning with Partial Participationcited by 3
  2. 2026Byzantine Machine Learning: MultiKrum and an optimal notion of robustnesscited by 1
  3. 2025GRANITE : a Byzantine-Resilient Dynamic Gossip Learning Frameworkcited by 0
  4. 2025Robust and Efficient Collaborative Learningcited by 0
  5. 2026Topology-Independent Robustness of the Weighted Mean under Label Poisoning Attacks in Heterogeneous Decentralized Learningcited by 0
  6. 2026From Inexact Gradients to Byzantine Robustness: Acceleration and Optimization under Similaritycited by 0
  7. 2026Decentralized Ranking Aggregation via Gossip: Convergence and Robustnesscited by 0
  8. 2026Optimizing Byzantine Node Placement in Decentralized Federated Learningcited by 0