Research map / Probabilistic and scientific modeling / Variational and generative sampling
Continuous-Time Flows for Efficient Inference and Density Estimation
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 |
|---|---|---|---|
| Distill the CTF into the inference network using Wasserstein adversarial distribution matching instead of standard Euclidean distance (ablation) | MacVAE inference · Synthetic 2D distributions | distribution matching quality | Helped |
| Use standard Euclidean distance for distribution matching instead of Wasserstein adversarial matching (ablation) | MacVAE inference · Synthetic 2D distributions | distribution matching quality | Hurt |
| Use standard Euclidean distance for distribution matching instead of Wasserstein adversarial matching (ablation) | MacVAE inference · Synthetic 4-point VAE dataset | posterior approximation quality | Hurt |
| Use Wasserstein adversarial distribution matching to distill the CTF (ablation) | MacVAE inference · Synthetic 4-point VAE dataset | posterior approximation quality | Helped |
| Distill the CTF into the inference network | MacVAE · MNIST | test ELBO | Helped |
| Use MacGAN instead of SteinGAN | MacGAN · CIFAR-10 | Inception score: 6.35 -> 6.49 | Helped |
| Use MacGAN instead of WGAN-I | MacGAN · CIFAR-10 | Inception score: 6.25 -> 6.49 | Helped |
| Use MacGAN instead of DCGAN | MacGAN · CIFAR-10 | Inception score: 6.58 -> 6.49 | Hurt |
| Vary the discretization stepsize across the tested interval (ablation) | MacGAN · MNIST | log-likelihood | No clear effect |
Closest papers by meaning
- Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Models (AAAI 2018 · similarity 0.89)
- Adaptive Density Estimation for Generative Models (NeurIPS 2019 · similarity 0.88)
- Flows for simultaneous manifold learning and density estimation (NeurIPS 2020 · similarity 0.87)
- Adversarial Flow Models (ICML 2026 · similarity 0.87)
- Analysis of Learning a Flow-based Generative Model from Limited Sample Complexity (ICLR 2024 · similarity 0.85)
- Normalizing Flows are Capable Generative Models (ICML 2025 · similarity 0.85)
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.