Research map / Optimization / Distributed and compressed
Compressed distributed optimization
Reduce communication in decentralized and distributed training via compression and error feedback.
30 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · 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.
10 papers in 2023–24 and 7 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 14.1% to 9.3%.
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
- 2016 · Bootstrap Model Aggregation for Distributed Statistical Learning (NeurIPS 2016 · 7 citations)
- 2017 · Collaborative Deep Learning in Fixed Topology Networks (NeurIPS 2017 · 115 citations)
- 2018 · Improving the Privacy and Accuracy of ADMM-Based Distributed Algorithms (ICML 2018 · 56 citations)
- 2020 · Communication-Efficient Distributed PCA by Riemannian Optimization (ICML 2020 · 6 citations)
- 2021 · Quasi-global Momentum: Accelerating Decentralized Deep Learning on Heterogeneous Data (ICML 2021 · 32 citations)
- 2022 · BEER: Fast $O(1/T)$ Rate for Decentralized Nonconvex Optimization with Communication Compression (NeurIPS 2022 · 6 citations)
- 2023 · Incentivizing Honesty among Competitors in Collaborative Learning and Optimization (NeurIPS 2023 · 4 citations)
- 2024 · DePRL: Achieving Linear Convergence Speedup in Personalized Decentralized Learning with Shared Representations (AAAI 2024 · 7 citations)
- 2026 · Distributed Direct Preference Optimization (ICML 2026 · 0 citations)
Earliest papers here
- Bootstrap Model Aggregation for Distributed Statistical Learning (NeurIPS 2016 · 7 citations)
- Collaborative Deep Learning in Fixed Topology Networks (NeurIPS 2017 · 115 citations)
- Improving the Privacy and Accuracy of ADMM-Based Distributed Algorithms (ICML 2018 · 56 citations)
What has been tried
22 papers here report 203 measured changes (a component added, removed or swapped, compared with a baseline): helped 57%, no clear effect 15%, hurt 24%, mixed 4%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| use constant pair-sampling rate O(1/n) rather than updating all pairs | CIFAR-100 cross-silo (8 clients), sampling strategies · accuracy | 74.93 -> 73.05 | Hurt | CoBo: Collaborative Learning via Bilevel Optimization (NeurIPS 2024 · 2 citations) |
| Add LASSO feature selection to the projection anomaly estimate | Gradient aggregation experiments · gradient aggregation performance | Hurt | Robust Distributed Gradient Aggregation Using Projections onto Gradient Manifolds (AAAI 2024 · 1 citation) | |
| add clipping to gradient differences | a9a · optimality gap | Helped | Byzantine Robustness and Partial Participation Can Be Achieved at Once: Just Clip Gradient Differences (NeurIPS 2024 · 0 citations) |
What reviewers will expect
Datasets and baselines named in the 10 papers of this approach from 2024–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- CIFAR-10 · 4 of 10 papers
- CIFAR-100 · 3 of 10 papers
- Fashion-MNIST · 2 of 10 papers
- TinyImageNet · 2 of 10 papers
- MNIST · 2 of 10 papers
- a9a · 2 of 10 papers
Baselines
- FedAvg · 3 of 10 papers
- Ditto · 2 of 10 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 17 papers; from OpenAlex and the papers' affiliations).
- Aalto University · 2 papers
- King Abdullah University of Science and Technology · 2 papers
- EPFL · 2 papers
Use it
Download BibTeX of 28 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)
- Byzantine-robust learning (32 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).