Research map / Probabilistic and scientific modeling / Variational and generative sampling
Variational Autoencoder with Implicit Optimal Priors
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use an implicit aggregated-posterior prior instead of the standard Gaussian prior | VAE · MNIST | test log-likelihood: -85.84 ± 0.07 -> ≈ -83.21 ± 0.13 | Helped |
| Use an implicit aggregated-posterior prior instead of the standard Gaussian prior | VAE · OMNIGLOT | test log-likelihood: -111.39 ± 0.11 -> ≈ -108.48 ± 0.16 | Helped |
| Use an implicit aggregated-posterior prior instead of the standard Gaussian prior | VAE · FreyFaces | test log-likelihood: 1382.53 ± 3.57 -> ≈ 1396.27 ± 2.75 | Helped |
| Use an implicit aggregated-posterior prior instead of the standard Gaussian prior | VAE · Histopathology | test log-likelihood: 1081.53 ± 0.70 -> ≈ 1087.42 ± 0.60 | Helped |
| Estimate the aggregated-posterior prior implicitly rather than using VampPrior | VAE · MNIST | test log-likelihood: -83.90 ± 0.08 -> ≈ -83.21 ± 0.13 | Helped |
| Estimate the aggregated-posterior prior implicitly rather than using VampPrior | VAE · OMNIGLOT | test log-likelihood: -110.53 ± 0.09 -> ≈ -108.48 ± 0.16 | Helped |
| Estimate the aggregated-posterior prior implicitly rather than using VampPrior | VAE · FreyFaces | test log-likelihood: 1392.62 ± 6.25 -> ≈ 1396.27 ± 2.75 | No clear effect |
| Estimate the aggregated-posterior prior implicitly rather than using VampPrior | VAE · Histopathology | test log-likelihood: 1083.11 ± 2.10 -> ≈ 1087.42 ± 0.60 | Helped |
| Use the proposed method instead of VampPrior | VAE on MNIST · MNIST training | training speed: 2.83 times faster | Helped |
| Vary the number of VampPrior mixture components | VAE with VampPrior · Histopathology | test log-likelihood | Mixed |
Closest papers by meaning
- MAE: Mutual Posterior-Divergence Regularization for Variational AutoEncoders (ICLR 2019 · similarity 0.93)
- Consistency Regularization for Variational Auto-Encoders (NeurIPS 2021 · similarity 0.92)
- From Variational to Deterministic Autoencoders (ICLR 2020 · similarity 0.92)
- Unscented Autoencoder (ICML 2023 · similarity 0.92)
- Beyond Vanilla Variational Autoencoders: Detecting Posterior Collapse in Conditional and Hierarchical Variational Autoencoders (ICLR 2024 · similarity 0.90)
- $t^3$-Variational Autoencoder: Learning Heavy-tailed Data with Student's t and Power Divergence (ICLR 2024 · similarity 0.90)
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.