Research map / Probabilistic and scientific modeling / Variational and generative sampling
Deep Generative Learning via Variational Gradient Flow
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 JS divergence instead of KL divergence (ablation) | VGrow · MNIST | FID: 3.66 (0.09) -> 3.32 (0.05) | Helped |
| Use JS divergence instead of KL divergence (ablation) | VGrow · FashionMNIST | FID: 9.30 (0.09) -> 8.75 (0.06) | Helped |
| Use JS divergence instead of KL divergence (ablation) | VGrow · CIFAR10 | FID: 29.7 (0.1) -> 29.1 (0.1) | Helped |
| Use logD divergence instead of KL divergence (ablation) | VGrow · MNIST | FID: 3.66 (0.09) -> 3.64 (0.05) | No clear effect |
| Use logD divergence instead of KL divergence (ablation) | VGrow · FashionMNIST | FID: 9.30 (0.09) -> 9.51 (0.09) | Hurt |
| Use logD divergence instead of KL divergence (ablation) | VGrow · CIFAR10 | FID: 29.7 (0.1) -> 28.8 (0.1) | Helped |
| Use Jeffrey divergence instead of KL divergence (ablation) | VGrow · MNIST | FID: 3.66 (0.09) -> 3.40 (0.07) | Helped |
| Use Jeffrey divergence instead of KL divergence (ablation) | VGrow · FashionMNIST | FID: 9.30 (0.09) -> 9.72 (0.06) | Hurt |
| Use Jeffrey divergence instead of KL divergence (ablation) | VGrow · CIFAR10 | FID: 29.7 (0.1) -> 32.3 (0.1) | Hurt |
| Use VGrow-logD instead of the strongest listed baseline SN-SWGAN | CIFAR10 · VGrow | FID: 28.5 (0.2) -> 28.8 (0.1) | Hurt |
Closest papers by meaning
- Refining Deep Generative Models via Discriminator Gradient Flow (ICLR 2021 · similarity 0.90)
- Deep MMD Gradient Flow without adversarial training (ICLR 2025 · similarity 0.88)
- Statistically Optimal Generative Modeling with Maximum Deviation from the Empirical Distribution (ICML 2024 · similarity 0.87)
- On Divergence Measures for Training GFlowNets (NeurIPS 2024 · similarity 0.87)
- A solvable model of learning generative diffusion: theory and insights (NeurIPS 2025 · similarity 0.87)
- Analysis of Learning a Flow-based Generative Model from Limited Sample Complexity (ICLR 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.