Research map / Probabilistic and scientific modeling / Variational and generative sampling
Large-Scale Stochastic Sampling from the Probability Simplex
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 |
|---|---|---|---|
| Use SCIR instead of SGRLD | Simplex sampling on sparse Dirichlet posterior · Sparse synthetic Dirichlet posterior | Kolmogorov-Smirnov distance | Helped |
| Use SCIR instead of SGRLD | Simplex sampling on dense Dirichlet posterior · Dense synthetic Dirichlet posterior | Kolmogorov-Smirnov distance | No clear effect |
| Use SCIR instead of SGRLD | LDA · Wikipedia documents | Perplexity | Helped |
| Use SCIR instead of SGRLD | Bayesian nonparametric mixture model · Anonymous Microsoft user dataset | Log predictive test score | Helped |
| Use stochastic-gradient samplers instead of the Gibbs slice sampler | Bayesian nonparametric mixture model · Anonymous Microsoft user dataset | Log predictive test score | Hurt |
| Use SCIR instead of SGRLD | Simplex sampling · Sparse Dirichlet posterior | Ability to sample small simplex components | Helped |
| Use SCIR instead of SGRLD | Simplex sampling · Sparse versus dense synthetic Dirichlet posteriors | Sampler performance | Mixed |
| Use a larger vocabulary | LDA with SCIR · Wikipedia documents | Perplexity | Mixed |
| Use a decreasing stepsize scheme (ablation) | SCIR · LDA | Perplexity | Helped |
| Increase the Dirichlet process concentration parameter (ablation) | SGRLD · Anonymous Microsoft user dataset | Log predictive test score | Hurt |
| Use SCIR without discretizing the underlying CIR process | SCIR · Gamma target | Asymptotic bias | Helped |
Closest papers by meaning
- Structured Stochastic Gradient MCMC (ICML 2022 · similarity 0.91)
- Stochastic Gradient Richardson-Romberg Markov Chain Monte Carlo (NeurIPS 2016 · similarity 0.90)
- Multi-Marginal Couplings for Metropolis--Hastings (NeurIPS 2026 · similarity 0.88)
- Parameter Expanded Stochastic Gradient Markov Chain Monte Carlo (ICLR 2025 · similarity 0.87)
- Parallelizing MCMC Across the Sequence Length (NeurIPS 2025 · similarity 0.86)
- Faster Sampling via Stochastic Gradient Proximal Sampler (ICML 2024 · similarity 0.86)
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.