AI Paper Atlas

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

Nonparametric Score Estimators

Yuhao Zhou, Jiaxin Shi, Jun Zhu

ICML 2020 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Score-based generative modeling analysis · 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 curl-free kernels instead of diagonal kernelsNonparametric score estimators · High-dimensional grid distributionsscore estimation accuracyImproved
Use curl-free kernels instead of diagonal kernelsWAE score estimators · MNIST and CelebA, high latent dimensionsnegative log-likelihood / FID: 92.05 vs 114.57; 50.66 vs 76.85Improved
Use curl-free kernels instead of diagonal kernelsNonparametric score estimators · Low-dimensional grid distributionsscore estimation accuracyNo clear effect
Use ν-method iterative regularizationCurl-free score estimation · MNIST WAE, latent dim 64negative log-likelihood: 91.48 vs 93.06Improved
Use ν-method iterative regularizationCurl-free score estimation · CelebA WAE, latent dim 64FID: 50.16 vs 72.68Improved
Use ν-method iterative regularizationCurl-free score estimation · MNIST WAE, latent dim 8 and 32negative log-likelihood: 97.28 vs 96.98; 90.94 vs 89.06Hurt
Use ν-method iterative regularizationCurl-free score estimation · CelebA WAE, latent dim 8 and 32FID: 72.02 vs 69.72; 52.86 vs 49.93Hurt
Approximate KEF with the conjugate gradient solverKEF · MNIST WAE, latent dim 128time per epoch: 201.1s -> 13.7sImproved
Replace full KEF with low-rank NKEFKEF · MNIST WAE, latent dims 32, 64, and 128negative log-likelihood: 90.93 -> 92.29; 91.58 -> 92.82; 92.40 -> 94.14Hurt
Increase the NKEF low-rank setting from NKEF₂ to NKEF₄ or NKEF₈NKEF · MNIST WAE, latent dim 32negative log-likelihood: 92.29 -> 91.19 -> 90.86Mixed
Increase the NKEF low-rank setting from NKEF₂ to NKEF₄NKEF · CelebA WAE, latent dims 32 and 64FID: 53.92 -> 54.54; 51.16 -> 50.76Mixed

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