AI Paper Atlas

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

Trust-Region Variational Inference with Gaussian Mixture Models

Oleg Arenz, Mingjun Zhong, Gerhard Neumann

JMLR 2020 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Annealing and divergence-based variational objectives · 6 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 Mahalanobis distance instead of KL divergence as dissimilarity measure for sample selection (ablation)VIPS++ · GMM (20D) and Goodwinsample efficiencyNo clear effect
Cycle the assumed initial weight Delta over several values instead of a fixed value (ablation)VIPS++ component adding heuristic · Planar robot (4 goals)learning curvesImproved
Add line-search interpolation between isotropic and averaged covariance for new components (ablation)VIPS · new experiment with covariance init failure caseperformanceImproved
Use VIPS1 instead of reparameterization trick for Gaussian updates (ablation)Gaussian variational approximation · Planar robot (1 goal) and breast cancerELBO learning curvesImproved
Add KL constraint between successive updates (ablation)VIPS1 · Planar robot (1 goal) and breast cancerELBO stabilityImproved
Sample reuse, line-search covariance init and adaptive KL bounds and regularization (VIPS++ over VIPS)VIPS · benchmark suitesample complexity: approximately one order of magnitudeImproved

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