AI Paper Atlas

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

Nested Variational Inference

Heiko Zimmermann, Hao Wu, Babak Esmaeili, Jan-Willem van de Meent

NeurIPS 2021 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Variational inference theory and gradients · 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
Use resampling and learn the annealing pathNVI annealing sampler · Multimodal density samplinglog average weightImproved
Use resampling and learn the annealing pathNVI annealing sampler · Multimodal density samplingeffective sample sizeImproved
Add importance resampling at each step (ablation)NVI annealing sampler · Multimodal density samplingeffective sample sizeImproved
Learn the annealing path instead of using a linear schedule (ablation)NVI annealing sampler · Multimodal density samplinglog average weightImproved
Use a linear annealing path instead of a learned path (ablation)NVI annealing sampler · Multimodal density samplinglog average weightHurt
Use a flow-based proposalNVI flow sampler · Multimodal density samplingeffective sample sizeImproved
Optimize both the forward and reverse densities rather than only the forward density (ablation)NVI heuristic-factor method · HMM future-observation heuristicslog average weightHurt
Learn a neural heuristic factorNVI heuristic-factor method · HMM future-observation heuristicslog average weightImproved
Use a neural heuristic instead of no heuristic (ablation)NVI heuristic-factor method · HMM future-observation heuristicslog average weightImproved
Train with NVI instead of RWSBGMM-VAE · MNIST and FashionMNISTposterior approximation qualityImproved

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