Research map / Adversarial robustness and security / Attacks on text and graphs
Learning from History for Byzantine Robust Optimization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace median/Krum/RFA aggregators with centered clipping | Byzantine-robust aggregation · Imbalanced MNIST (gamma=0.5) | test accuracy: 75% cap for median-based rules vs no such bound | Helped |
| add worker momentum to robust aggregators (ablation) | CM, RFA, TM, Krum, CC · CIFAR-10 ResNet-20 under BF, LF, ALIE, IPM attacks | test accuracy | Helped |
| use centered clipping instead of other aggregators with momentum | Byzantine-robust SGD with momentum · CIFAR-10 ResNet-20 | test accuracy | Helped |
| vary clipping radius tau and iterations l of centered clipping (ablation) | Centered clipping · CIFAR-10 ResNet-20, momentum 0 | final test accuracy | No clear effect |
| use history-oblivious (permutation-invariant) aggregators under time-coupled IPM attack | CM, RFA, Krum, TM · CIFAR-10 | test accuracy | Hurt |
Closest papers by meaning
- On the Tension between Byzantine Robustness and No-Attack Accuracy in Distributed Learning (ICML 2025 · similarity 0.93)
- From Inexact Gradients to Byzantine Robustness: Acceleration and Optimization under Similarity (NeurIPS 2026 · similarity 0.92)
- Byzantine-Robust Decentralized Learning via Remove-then-Clip Aggregation (AAAI 2024 · similarity 0.92)
- Byzantine Robustness and Partial Participation Can Be Achieved at Once: Just Clip Gradient Differences (NeurIPS 2024 · similarity 0.90)
- Fault Tolerant ML: Efficient Meta-Aggregation and Synchronous Training (ICML 2024 · similarity 0.90)
- Dynamic Byzantine-Robust Learning: Adapting to Switching Byzantine Workers (ICML 2024 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.