Research map / Probabilistic and scientific modeling / Variational and generative sampling
An Information-Theoretic Evaluation of Generative Models in Learning Multi-modal Distributions
Mohammad Jalali, Cheuk Ting Li, Farzan Farnia
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| compare RKE-MC against dataset diversity | generative models (GANs, VAEs, diffusion) · CIFAR-10 / ImageNet | RKE-MC | Hurt |
| use StyleGAN-XL instead of BigGAN-256 | class-conditional GANs · ImageNet | RKE-MC: 106.07 -> 1375.17 | Improved |
| use ADM-G diffusion instead of SAGAN-256 | generative model · ImageNet | RKE-MC: 105.57 -> 1407.75 | Improved |
| use ContraGAN instead of SAGAN | GAN · ImageNet (256) | RKE-MC: 105.57 -> 152.89 | Improved |
| use ContraGAN instead of SAGAN | GAN · ImageNet (256) | Coverage: 0.35 -> 0.33 | Hurt |
| use WGAN-WC instead of DCGAN | GAN · CIFAR-10 | Inception Score: 5.75 -> 2.59 | Hurt |
| use WGAN-WC instead of DCGAN | GAN · CIFAR-10 | RKE-MC: 10.19 -> 10.64 | No clear effect |
| use GigaGAN instead of BK-SDM (Base) | text-to-image model · MS-COCO zero-shot 30K | RRKE: 0.44 -> 0.39 | Improved |
| use BK-SDM (Base) instead of GigaGAN | text-to-image model · MS-COCO zero-shot 30K | RKE-MC: 58.78 -> 73.05 | Improved |
| increase truncation factor psi (ablation) | StyleGAN · AFHQ and FFHQ | RKE-MC | Improved |
| increase number of samples for RKE-MC estimation (ablation) | RKE-MC · ImageNet | RKE-MC | No clear effect |
Closest papers by meaning
- An Interpretable Evaluation of Entropy-based Novelty of Generative Models (ICML 2024 · similarity 0.89)
- Towards a Scalable Reference-Free Evaluation of Generative Models (NeurIPS 2024 · similarity 0.88)
- A Unifying Information-theoretic Perspective on Evaluating Generative Models (AAAI 2025 · similarity 0.88)
- Image Generation Diversity Issues and How to Tame Them (CVPR 2025 · similarity 0.86)
- PQMass: Probabilistic Assessment of the Quality of Generative Models using Probability Mass Estimation (ICLR 2025 · similarity 0.86)
- Sampling Multimodal Distributions with the Vanilla Score: Benefits of Data-Based Initialization (ICLR 2024 · similarity 0.84)
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.
