Research map / Optimization / Optimizers
A Contour Stochastic Gradient Langevin Dynamics Algorithm for Simulations of Multi-modal Distributions
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 |
|---|---|---|---|
| add contour (flattened-density importance sampling) to SGHMC | SGHMC · CIFAR10 | accuracy BMA: 91.98 -> 92.85 | Helped |
| add contour sampling to SGHMC | SGHMC · CIFAR100 | accuracy BPE: 61.46 -> 63.97 | Helped |
| add contour sampling to SGHMC | SGHMC · CIFAR100 | accuracy BMA: 66.43 -> 68.94 | Helped |
| add contour sampling to SGHMC with simulated annealing | saSGHMC · CIFAR10 | accuracy BPE: 93.80 -> 94.06 | Helped |
| add contour sampling to SGHMC with simulated annealing | saSGHMC · CIFAR100 | accuracy BPE: 69.24 -> 70.18 | Helped |
| add contour sampling to SGLD | SGLD · UCI regression (4 datasets) | RMSE: 4.12±0.10 -> 3.98±0.11 (Concrete) | Helped |
| add contour sampling to SGHMC | SGHMC · UCI Yacht | RMSE: 0.66±0.07 -> 0.72±0.09 | Hurt |
| sample from known flattened density (KSGLD) vs plain SGLD (ablation) | SGLD · Gaussian mixture | posterior mean estimation error | Helped |
| adaptive estimation of theta (CSGLD) vs known theta (KSGLD) (ablation) | CSGLD · Gaussian mixture | estimation error | No clear effect |
Closest papers by meaning
- Interacting Contour Stochastic Gradient Langevin Dynamics (ICLR 2022 · similarity 0.93)
- The promises and pitfalls of Stochastic Gradient Langevin Dynamics (NeurIPS 2018 · similarity 0.92)
- Preconditioned Stochastic Gradient Langevin Dynamics for Deep Neural Networks (AAAI 2016 · similarity 0.90)
- Variance Reduction in Stochastic Gradient Langevin Dynamics (NeurIPS 2016 · similarity 0.90)
- Parameter Expanded Stochastic Gradient Markov Chain Monte Carlo (ICLR 2025 · similarity 0.88)
- Generalization of noisy SGD in unbounded non-convex settings (ICML 2025 · similarity 0.88)
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.