AI Paper Atlas

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

Exemplar VAE: Linking Generative Models, Nearest Neighbor Retrieval, and Data Augmentation

Sajad Norouzi, David J Fleet, Mohammad Norouzi

NeurIPS 2020 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Structured latent VAEs · 7 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 an exemplar-based latent mixture prior instead of a Gaussian priorVAE · Dynamic MNISTtest log-likelihood lower bound (nats): -84.45 ± 0.12 -> -82.09 ± 0.18Improved
Use an exemplar-based latent mixture prior instead of a Gaussian priorVAE · Fashion MNISTtest log-likelihood lower bound (nats): -228.70 ± 0.15 -> -226.75 ± 0.07Improved
Use an exemplar-based latent mixture prior instead of a Gaussian priorVAE · Omniglottest log-likelihood lower bound (nats): -108.34 ± 0.06 -> -105.22 ± 0.18Improved
Use an exemplar-based latent mixture prior instead of a Gaussian priorPixelSNAIL · Dynamic MNISTtest log-likelihood lower bound (nats): -78.20 ± 0.02 -> -77.95 { ± 0.01}Hurt
Add Exemplar VAE-generated examples to label-smoothed classifier trainingMLP classifier · Permutation-invariant MNIST, 2×8192 hidden unitstest error (%): 1.17 -> 0.69Improved
Add Exemplar VAE-generated examples to label-smoothed classifier trainingMLP classifier · Permutation-invariant Fashion MNIST, 2×8192 hidden unitstest error (%): 8.56 -> 8.16Improved
Add leave-one-out exemplar training (ablation)Exemplar VAE · MNISTtest log-likelihood lower bound (nats): -101.33 -> -82.35Improved
Add leave-one-out exemplar training (ablation)Exemplar VAE · Omniglottest log-likelihood lower bound (nats): -139.12 -> -105.80Improved
Subsample exemplars to M/N=0.5 rather than using M/N=1 (ablation)Exemplar VAE with LOO · MNISTtest log-likelihood lower bound (nats): -82.35 -> -82.09Improved
Subsample exemplars to M/N=0.2 rather than using M/N=1 (ablation)Exemplar VAE with LOO · Omniglottest log-likelihood lower bound (nats): -105.80 -> -104.95Improved

Later papers that compared with this method

1 later accepted paper reports Exemplar VAE (25000) as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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