AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Self-Adversarially Learned Bayesian Sampling

Yang Zhao, Jianyi Zhang, Changyou Chen

AAAI 2019 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Posterior sampling with diffusion guidance · 5 citations (OpenAlex)

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 changedWhereMeasuredEffect
use SAL-MC without an MH step instead of A-NICE-MC with MHBayesian sampling · Ring, Mog2, Mog6ESS/s: 121138 vs 43887; 72691 vs 17216; 65287 vs 11745Improved
use SAL-MC with MH instead of A-NICE-MC with MHBayesian sampling · Mog2, Mog6, Ring5ESS: 1172 vs 951; 978 vs 889; 414 vs 335Improved
add an MH step (ablation)SAL-MC · Ring, Mog2, Mog6ESS/s: 121138 -> 25624; 72691 -> 17531; 65287 -> 12235Hurt
add an MH step (ablation)SAL-MC · Heart, Australian, GermanESS: 1683 -> 1; 1385 -> 1; 1897 -> 1Hurt
use SAL-MC with MH instead of A-NICE-MC with MHBayesian logistic regression sampling · Heart, Australian, GermanESS: 1683 vs 663; 1385 vs 596; 1897 vs 483Improved
use SAL-MC instead of A-NICE-MCBayesian logistic regression · Heart, Australian, Germantest accuracy: 84.10%, 88.38%, 80.32%; the same as A-NICE-MCNo clear effect
inject Gaussian noise with variance 0.5 instead of using zero noise (ablation)learned generator sampling · MNISTmixing across digit modesImproved
inject noise that is too small or too large (ablation)SAL-MC · synthetic multimodal distributionseffective sample sizeHurt
use AG-SVGD instead of SGLDdirect Bayesian sampling · 5D Gaussian mixturevariance estimation MSEImproved
use AG-SVGD instead of A-SVGDdirect Bayesian sampling · 5D Gaussian mixturesample approximation qualityNo clear effect
use AG-SVGD instead of A-SVGD or SGLDdirect Bayesian sampling · 2D-GMM Mog4mode balanceImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy