Research map / Probabilistic and scientific modeling / Variational and generative sampling
On Implicit Regularization in $β$-VAEs
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 a diagonal mean-field covariance | VAE with Gaussian prior · Synthetic 2D latent data | uniqueness of inferred latents | Helped |
| Use block-diagonal variational covariance | β-VAE · MNIST | alignment of covariance precision and decoder Jacobian structure | Helped |
| Replace stochastic β-VAE training with the GRAE deterministic objective | β-VAE · MNIST | objective value agreement | Mixed |
| Replace β-VAE training with GRAE₍≈₎ training | β-VAE · CelebA | sample quality | No clear effect |
| Increase β | β-VAE · CelebA | sample sharpness | Hurt |
| Increase β | β-VAE Taylor approximation · MNIST | agreement with β-VAE objective | Hurt |
| Increase β | GRAE objective · MNIST | agreement with Taylor-approximated β-VAE objective | Hurt |
| Restrict variational covariance to a diagonal structure | β-VAE decoder · Gaussian observations | decoder Jacobian column orthogonality | Helped |
Closest papers by meaning
- From Variational to Deterministic Autoencoders (ICLR 2020 · similarity 0.92)
- InfoVAE: Balancing Learning and Inference in Variational Autoencoders (AAAI 2019 · similarity 0.92)
- Consistency Regularization for Variational Auto-Encoders (NeurIPS 2021 · similarity 0.91)
- Beyond Vanilla Variational Autoencoders: Detecting Posterior Collapse in Conditional and Hierarchical Variational Autoencoders (ICLR 2024 · similarity 0.89)
- Be a Goldfish: Forgetting Bad Conditioning in Sparse Linear Regression via Variational Autoencoders (ICML 2025 · similarity 0.89)
- Information-theoretic Generalization Analysis for VQ-VAEs: A Role of Latent Variables (NeurIPS 2025 · similarity 0.89)
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.