// ⛨ Safety, Evaluation & Security
Differentially Private SGD (DP-SGD)
Clips per-example gradients and adds noise to protect individuals.
差分隐私 SGD 2016 Advanced extended
IN PLAIN WORDS
Clip each example's gradient, add Gaussian noise to the sum, and track privacy budget with the moments accountant, so no single example is memorized.
Used in: Training on medical, financial or keyboard data
01Original paper
Deep Learning with Differential Privacy
Martín Abadi, Andy Chu, Ian Goodfellow, et al. · 2016
arXiv ↗ ✓ Verified · arXiv API
02Where it sits
Learn the left column first; the right column builds on this one.
Learn first
Differentially Private SGD (DP-SGD)
Then you can learn
- Nothing in the atlas builds on it yet
Related:LLM WatermarkingMembership Inference Attacks
Left: prerequisites · right: builds on this · dashed: related. Full map →
03Learning resources
Official docs, explainers, courses and reference code. All links checked.
Official docs
04Where the course teaches it
For AI / data
- Verification: verified, checked 2026-10-21.
- JSON: /api/learn/algorithms.json (id:
dp-sgd)