Research map / Optimization / Optimizers
Characterizing Membership Privacy in Stochastic Gradient Langevin Dynamics
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 |
|---|---|---|---|
| Train with SGLD instead of SGD | Three-layer fully connected network · German Credit | AUC: 0.677 -> 0.536 | Helped |
| Train with SGLD instead of SGD | ResNet-18 · IDC | AUC: 0.669 -> 0.641 | Helped |
| Ensemble SGLD models (ablation) | SGLD · German Credit | Test accuracy: 0.736 -> 0.751 | Helped |
| Ensemble SGLD models (ablation) | SGLD · German Credit | F1 score: 0.598 -> 0.564 | Helped |
| Ensemble SGLD models (ablation) | SGLD · IDC | Test accuracy: 0.817 -> 0.825 | Helped |
| Use a parameter-wise preconditioner (pSGLD) (ablation) | SGLD · German Credit | Test accuracy: 0.736 -> 0.753 | Helped |
| Use a parameter-wise preconditioner (pSGLD) (ablation) | SGLD · German Credit | AUC: 0.536 -> 0.551 | Hurt |
| Add Dropout | Three-layer fully connected network · German Credit | AUC: 0.677 -> 0.527 | Helped |
| Add Dropout | ResNet-18 · IDC | Test accuracy: 0.823 -> 0.818 | Hurt |
Closest papers by meaning
- The promises and pitfalls of Stochastic Gradient Langevin Dynamics (NeurIPS 2018 · similarity 0.91)
- Analyzing the Generalization Capability of SGLD Using Properties of Gaussian Channels (NeurIPS 2021 · similarity 0.90)
- Privacy of Noisy Stochastic Gradient Descent: More Iterations without More Privacy Loss (NeurIPS 2022 · similarity 0.88)
- Generalization of noisy SGD in unbounded non-convex settings (ICML 2025 · similarity 0.85)
- Rethinking DP-SGD in Discrete Domain: Exploring Logistic Distribution in the Realm of signSGD (ICML 2024 · similarity 0.85)
- Differentially Private Stochastic Gradient Descent with Fixed-Size Minibatches: Tighter RDP Guarantees with or without Replacement (NeurIPS 2024 · similarity 0.85)
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.