Research map / Probabilistic and scientific modeling / Variational and generative sampling
Herded Gibbs Sampling
Yutian Chen, Luke Bornn, Nando de Freitas, Mareija Eskelin, Jing Fang, Max Welling
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 random Gibbs sampling with herded Gibbs | Gibbs sampler · Ising denoising sigma=8 | reconstruction error (x1e-3): 90.27 -> 67.93 | Improved |
| replace Gibbs with herded Gibbs | Gibbs sampler · Ising denoising sigma=4 | reconstruction error (x1e-3): 37.20 -> 32.07 | Improved |
| replace Gibbs with herded Gibbs | Gibbs sampler · Ising denoising sigma=2 | reconstruction error (x1e-3): 21.63 -> 21.58 | No clear effect |
| use shared weights over equal-sum Markov blanket states | Herded Gibbs · Ising denoising sigma=8 | reconstruction error (x1e-3): 67.93 -> 58.49 | Improved |
| use shared weights | Herded Gibbs · Ising denoising sigma=2 | reconstruction error (x1e-3): 21.58 -> 22.24 | Hurt |
| use herded Gibbs instead of mean field (D=0.5) | Mean field · Ising denoising sigma=2 | reconstruction error (x1e-3): 15.52 -> 21.58 | Hurt |
| use herded Gibbs instead of mean field (D=0.5) | Mean field · Ising denoising sigma=8 | reconstruction error (x1e-3): 104.08 -> 67.93 | Improved |
| replace annealed Gibbs with herded Gibbs | Annealed Gibbs · NER CRF, 100 iterations | F1: 84.36 -> 84.70 | Improved |
| replace annealed Gibbs with herded Gibbs | Annealed Gibbs · NER CRF, 800 iterations | F1: 84.61 -> 84.81 | Improved |
| replace random number generator with CUD sequence (MCQMC Gibbs) | Gibbs · two-variable complete graph | L1 error of joint | No clear effect |
| replace Gibbs with herded Gibbs | Gibbs · two-variable complete graph | L1 error of joint | Improved |
| apply herded Gibbs to an incomplete four-variable graph | Herded Gibbs · four-variable incomplete graph | L1 error of joint | Hurt |
Closest papers by meaning
- Minibatch Gibbs Sampling on Large Graphical Models (ICML 2018 · similarity 0.88)
- On the properties of variational approximations of Gibbs posteriors (JMLR 2016 · similarity 0.87)
- Poisson-Minibatching for Gibbs Sampling with Convergence Rate Guarantees (NeurIPS 2019 · similarity 0.86)
- Amortized Population Gibbs Samplers with Neural Sufficient Statistics (ICML 2020 · similarity 0.86)
- Closed-Form Gibbs Sampling for Graphical Models with Algebraic Constraints (AAAI 2016 · similarity 0.86)
- Uncertainty Herding: One Active Learning Method for All Label Budgets (ICLR 2025 · similarity 0.84)
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.
