AI Paper Atlas

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

Advances in Black-Box VI: Normalizing Flows, Importance Weighting, and Optimization

Abhinav Agrawal, Daniel R. Sheldon, Justin Domke

NeurIPS 2020 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Normalizing flow training and design · 1 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
replace ADVI step scheme with comprehensive step searchfull-rank Gaussian BBVI · 30 Stan modelsELBO: 10 nats or more on a small fractionImproved
use STL gradient instead of closed-form entropyfull-rank Gaussian BBVI · 30 Stan modelsELBOImproved
add Laplace initializationfull-rank Gaussian BBVI · 30 Stan modelsELBOMixed
add importance-weighted sampling, M=10Gaussian VI · 30 Stan modelsELBO: 1 nat or more on 30% of modelsImproved
add importance-weighted training with DReGGaussian VI · 30 Stan modelsELBO: 1 nat of improvement to almost 30%Mixed
use real-NVP normalizing flow instead of GaussianBBVI · 30 Stan modelsELBO: 1 nat or more on around half of modelsImproved
use STL estimator instead of middle estimator for flowsreal-NVP flow · 30 Stan modelsELBOMixed
add importance-weighted sampling to flowsreal-NVP flow · 30 Stan modelsELBOImproved
add importance-weighted training to flowsreal-NVP flow · 30 Stan modelsELBO: 1 nat or more for almost 10%Mixed
use full-rank instead of diagonal Gaussian (ablation)Gaussian VI · 30 Stan modelsELBO: 1 nats or more on at least halfImproved
use DReG instead of standard IW-ELBO gradientIW-training · 30 Stan modelsELBOImproved
full path of beneficial components versus ADVIbest method (flow + STL + IW sampling) · 30 Stan modelsELBO: 1 nat or more on at least 50%Improved

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