Research map / Optimization / Distributed and compressed
Byzantine-robust learning
Defend distributed training against adversarial workers with robust aggregation.
32 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2017–2026) · an approach in Distributed and compressed, keyword cluster “decentralized · asynchronous · byzantine”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
6 papers in 2023–24 and 8 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 8.5% to 10.7%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2017 · Machine Learning with Adversaries: Byzantine Tolerant Gradient Descent (NeurIPS 2017 · 967 citations)
- 2018 · Byzantine-Robust Distributed Learning: Towards Optimal Statistical Rates (ICML 2018 · 612 citations)
- 2019 · RSA: Byzantine-Robust Stochastic Aggregation Methods for Distributed Learning from Heterogeneous Datasets (AAAI 2019 · 377 citations)
- 2020 · Election Coding for Distributed Learning: Protecting SignSGD against Byzantine Attacks (NeurIPS 2020 · 25 citations)
- 2021 · Distributed Momentum for Byzantine-resilient Stochastic Gradient Descent (ICLR 2021 · 21 citations)
- 2022 · Byzantine Machine Learning Made Easy By Resilient Averaging of Momentums (ICML 2022 · 8 citations)
- 2023 · Robust Collaborative Learning with Linear Gradient Overhead (ICML 2023 · 1 citation)
- 2024 · Byzantine-Robust Decentralized Learning via Remove-then-Clip Aggregation (AAAI 2024 · 9 citations)
- 2025 · Adaptive Gradient Clipping for Robust Federated Learning (ICLR 2025 · 0 citations)
- 2026 · Improving the Robustness-Utility Trade-off in Decentralized Learning over Sparse Networks (ICML 2026 · 0 citations)
Earliest papers here
- Machine Learning with Adversaries: Byzantine Tolerant Gradient Descent (NeurIPS 2017 · 967 citations)
- Byzantine-Robust Distributed Learning: Towards Optimal Statistical Rates (ICML 2018 · 612 citations)
- The Hidden Vulnerability of Distributed Learning in Byzantium (ICML 2018 · 557 citations)
What has been tried
25 papers here report 203 measured changes (a component added, removed or swapped, compared with a baseline): helped 53%, no clear effect 13%, hurt 27%, mixed 7%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Add ω-CTMA to weighted robust aggregation | CIFAR-10, Byzantine scenarios · test accuracy | Helped | Weight for Robustness: A Comprehensive Approach towards Optimal Fault-Tolerant Asynchronous ML (NeurIPS 2024 · 1 citation) | |
| Degrade initialization from μ=1 toward μ=5 while using ARC-enhanced Robust-DSGD | MNIST, extreme heterogeneity, α = 0.1, f = 1 · accuracy | Hurt | Adaptive Gradient Clipping for Robust Federated Learning (ICLR 2025 · 0 citations) | |
| Add nearest neighbour mixing (NNM) with fNNM = 7 to Multi-Krum (f = 7) | CIFAR-10 test, α = 0.1 · top-1 test accuracy | 40.31% -> 75.03% | Helped | On the Tension between Byzantine Robustness and No-Attack Accuracy in Distributed Learning (ICML 2025) |
What reviewers will expect
Datasets and baselines named in the 6 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- MNIST · 4 of 6 papers
- CIFAR-10 · 3 of 6 papers
Baselines
- D-GD · 2 of 6 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 14 papers; from OpenAlex and the papers' affiliations).
- Technion - Israel Institute of Technology · 3 papers
- Nanjing University · 2 papers
Use it
Download BibTeX of 31 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Decentralized optimization algorithms (94 papers)
- Local and asynchronous SGD (84 papers)
- Compressed distributed optimization (30 papers)
Elsewhere in the topic
- Gradient quantization and sparsification (57 papers)
- Communication complexity theory (35 papers)
- Error feedback theory (17 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).