Research map / Probabilistic and scientific modeling / Variational and generative sampling
Adversarial Symmetric Variational Autoencoder
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 AS-VAE instead of NF (k=80) | MNIST · MNIST | NLL (nats): 85.1 -> 82.51 | Helped |
| Use AS-VAE instead of PixelRNN | MNIST · MNIST | NLL (nats): 79.2 -> 82.51 | Hurt |
| Train with only the data-side objective (ablation) | AS-VAE · MNIST | NLL (nats): 82.51 -> 81.14 | Helped |
| Train with only the code-side objective (ablation) | AS-VAE · MNIST | NLL (nats): 82.51 -> 146.32 | Hurt |
| Use the full symmetric objective rather than only the data-side objective (ablation) | AS-VAE · CIFAR-10 | IS: 2.91 -> 6.34 | Helped |
| Use the full symmetric objective rather than only the data-side objective (ablation) | AS-VAE · CIFAR-10 | NLL(bits): 3.09 -> 3.32 | Hurt |
| Use the full symmetric objective rather than only the code-side objective (ablation) | AS-VAE · CIFAR-10 | IS: 6.89 -> 6.34 | Hurt |
| Use the full symmetric objective rather than only the code-side objective (ablation) | AS-VAE · CIFAR-10 | RMSE: 13.12 -> 3.36 | Helped |
| Use AS-VAE instead of PixelCNN++ | ImageNet 2012 (64 × 64) · ImageNet 2012 (64 × 64) | IS: 7.65 -> 11.14 | Helped |
| Use AS-VAE instead of PixelCNN++ | ImageNet 2012 (64 × 64) · ImageNet 2012 (64 × 64) | NLL: 3.27 -> 3.71 | Hurt |
Closest papers by meaning
- Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks (ICML 2017 · similarity 0.91)
- MAE: Mutual Posterior-Divergence Regularization for Variational AutoEncoders (ICLR 2019 · similarity 0.91)
- Consistency Regularization for Variational Auto-Encoders (NeurIPS 2021 · similarity 0.91)
- From Variational to Deterministic Autoencoders (ICLR 2020 · similarity 0.91)
- Marginalization is not Marginal: No Bad VAE Local Minima when Learning Optimal Sparse Representations (ICML 2023 · similarity 0.91)
- Variational Autoencoder with Implicit Optimal Priors (AAAI 2019 · similarity 0.91)
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.