Research map / Optimization / Optimizers
Variance Reduction 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 |
|---|---|---|---|
| Replace SGLD stochastic gradient with SAGA-LD variance-reduced gradient estimator (stored per-datapoint gradients) | Stochastic gradient Langevin dynamics · UCI regression datasets (concrete, noise, parkinsons, toms, 3dRoad) | Test MSE vs. number of passes through data | Helped |
| Replace SGLD with SAGA-LD | Stochastic gradient Langevin dynamics · UCI classification datasets (pima, diabetic, eeg, space, susy) | Test set log-likelihood vs. passes through data | Helped |
| Replace SGLD with SAGA-LD | Stochastic gradient Langevin dynamics · Bayesian ICA (MEG data, 10 channels) | Test set log-likelihood | Helped |
| Replace SGLD with SAGA-LD: lower gradient variance with respect to the true gradient | Stochastic gradient Langevin dynamics · Regression (concrete) and classification (pima) | Variance of stochastic gradient estimate | Helped |
| Replace SGLD with SAGA-LD in mixture model posterior estimation | Stochastic gradient Langevin dynamics · Two-Gaussian mixture (20,000 points) | Estimated vs true posterior over mu | No clear effect |
| Use SVRG-LD (memory-lean variant, full gradient every m iterations) instead of SAGA-LD | Stochastic gradient Langevin dynamics · Experiments overall | Empirical performance | No clear effect |
Closest papers by meaning
- Improved Convergence Rate of Stochastic Gradient Langevin Dynamics with Variance Reduction and its Application to Optimization (NeurIPS 2022 · similarity 0.93)
- On the Theory of Variance Reduction for Stochastic Gradient Monte Carlo (ICML 2018 · similarity 0.92)
- Accurate Large-sample Uncertainty Quantification using Stochastic Gradient Markov Chain Monte Carlo (ICML 2026 · similarity 0.92)
- The promises and pitfalls of Stochastic Gradient Langevin Dynamics (NeurIPS 2018 · similarity 0.92)
- Stochastic Gradient Hamiltonian Monte Carlo Methods with Recursive Variance Reduction (NeurIPS 2019 · similarity 0.91)
- Faster Sampling via Stochastic Gradient Proximal Sampler (ICML 2024 · similarity 0.89)
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.