Research map / Probabilistic and scientific modeling / Variational and generative sampling
Exploiting Chain Rule and Bayes' Theorem to Compare Probability Distributions
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 |
|---|---|---|---|
| Replace the original GAN criterion with the CT cost | DCGAN · CIFAR-10 | FID: 30.2 -> 22.1 | Helped |
| Replace the original criterion with the CT cost | SNGAN · CelebA | FID: 21.7 -> 9.2 | Helped |
| Replace the original criterion with the CT cost | StyleGAN2 · CIFAR-10 | FID: 5.8 -> 2.9 | Helped |
| Replace the original criterion with the CT cost | MMD-GAN · LSUN-bedroom | FID: 32.0 -> 38.3 | Hurt |
| Replace the original criterion with the CT cost | StyleGAN2 · LSUN-bedroom | FID: 2.9 -> 6.3 | Hurt |
| Add CT cost to base models | DCGAN, SWG, MMD-GAN, SNGAN, StyleGAN2 · CIFAR-10 | Inception Score: 6.2 -> 7.5 | Helped |
| Use adversarially trained CT feature encoder instead of a discriminator-trained encoder (ablation) | CT · CIFAR-10 | FID: 29.7 -> 22.1 | Helped |
| Use adversarially trained CT encoder instead of a sliced Wasserstein encoder (ablation) | CT · CIFAR-10 | FID: 32.4 -> 22.1 | Helped |
| Use CT as the generator loss instead of MMD (rational quadratic kernel, MMD critic) (ablation) | MMD-GAN · CIFAR-10 | FID: 39.9 -> 24.1 | Helped |
| Train critic with CT instead of MMD (rational quadratic kernel, MMD generator loss) (ablation) | MMD-GAN · CIFAR-10 | FID: 39.9 -> 41.4 | No clear effect |
| Train critic with CT instead of MMD (distance kernel, MMD generator loss) (ablation) | MMD-GAN · CIFAR-10 | FID: 40.3 -> 30.9 | Helped |
| Use only forward CT (rho=1) instead of rho=0.5 (ablation) | CT-DCGAN · CIFAR-10 | FID: 22.1 -> 25.1 | Hurt |
Closest papers by meaning
- Dynamic Conditional Optimal Transport through Simulation-Free Flows (NeurIPS 2024 · similarity 0.88)
- Generative Conditional Distributions by Neural (Entropic) Optimal Transport (ICML 2024 · similarity 0.87)
- Analyzing and Improving Optimal-Transport-based Adversarial Networks (ICLR 2024 · similarity 0.86)
- A Theory for Conditional Generative Modeling on Multiple Data Sources (ICML 2025 · similarity 0.86)
- A Likelihood Based Approach to Distribution Regression Using Conditional Deep Generative Models (ICML 2025 · similarity 0.86)
- Statistically Optimal Generative Modeling with Maximum Deviation from the Empirical Distribution (ICML 2024 · similarity 0.86)
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.