Research map / Probabilistic and scientific modeling / Variational and generative sampling
Revisiting Structured Variational Autoencoders
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 structured prior-aware inference instead of RNN-MF | SVAE · LDS 32d | ELBO: -4.317 -> -2.599 | Helped |
| Use structured prior-aware inference instead of RNN-AR-L | SVAE · LDS 64d | ELBO: -4.665 -> -3.419 | Helped |
| Apply self-supervised masking (ablation) | SVAE · LDS, 5-dimensional latent space | Validation ELBO: -0.2763 -> -0.3015 | Hurt |
| Apply self-supervised masking (ablation) | SVAE · LDS, 5-dimensional latent space | Prediction: -0.1910 -> -0.1558 | Helped |
| Apply self-supervised masking (ablation) | RNN-AR-L · LDS, 5-dimensional latent space | Validation ELBO: -0.2757 -> -0.2886 | Hurt |
| Apply self-supervised masking (ablation) | RNN-AR-L · LDS, 5-dimensional latent space | Prediction: -0.1862 -> -0.1717 | Helped |
| Apply self-supervised masking (ablation) | RNN-MF · LDS, 5-dimensional latent space | Prediction: -0.1899 -> -0.1815 | Helped |
| Apply self-supervised masking (ablation) | RNN-AR-NL · LDS, 5-dimensional latent space | Prediction: -0.1798 -> -0.1783 | Helped |
| Apply self-supervised masking (ablation) | CNN-AR-L · LDS, 5-dimensional latent space | Prediction: -0.1861 -> -0.1719 | Helped |
Closest papers by meaning
- Unbiased learning of deep generative models with structured discrete representations (NeurIPS 2023 · similarity 0.94)
- Variational Autoencoders with Jointly Optimized Latent Dependency Structure (ICLR 2019 · similarity 0.91)
- Ladder Variational Autoencoders (NeurIPS 2016 · similarity 0.88)
- Variational Learning of Fractional Posteriors (ICML 2025 · similarity 0.88)
- Beyond Vanilla Variational Autoencoders: Detecting Posterior Collapse in Conditional and Hierarchical Variational Autoencoders (ICLR 2024 · similarity 0.87)
- Sparse Autoencoders, Again? (ICML 2025 · similarity 0.87)
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.