Research map / Image generation / GANs
HyperGAN: A Generative Model for Diverse, Performant Neural Networks
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 |
|---|---|---|---|
| Increase the HyperGAN ensemble from 1 to 100 networks | HyperGAN · MNIST | accuracy: 98.64 \raisebox{.2ex}{$\scriptstyle\pm$} .3 -> 99.31 \raisebox{.2ex}{$\scriptstyle\pm$} .02 | Helped |
| Increase the HyperGAN ensemble from 1 to 100 networks | HyperGAN · MNIST 5000 | accuracy: 96.69 \raisebox{.2ex}{$\scriptstyle\pm$} .3 -> 97.71 \raisebox{.2ex}{$\scriptstyle\pm$} .05 | Helped |
| Increase the HyperGAN ensemble from 1 to 100 networks | HyperGAN · CIFAR-5 | accuracy: 84.50 \raisebox{.2ex}{$\scriptstyle\pm$} .6 -> 85.81 \raisebox{.2ex}{$\scriptstyle\pm$} .02 | Helped |
| Increase the HyperGAN ensemble from 1 to 100 networks | HyperGAN · CIFAR-10 | accuracy: 76.32 \raisebox{.2ex}{$\scriptstyle\pm$} .3 -> 77.71 \raisebox{.2ex}{$\scriptstyle\pm$} .03 | Helped |
| Remove latent-space prior matching with the discriminator (ablation) | HyperGAN · CIFAR-10 | accuracy | No clear effect |
| Add latent-space prior matching with the discriminator (ablation) | HyperGAN · CIFAR-10 | parameter diversity | Helped |
| Remove the mixer and use independent isotropic Gaussian inputs for each layer generator (ablation) | HyperGAN · CIFAR-10 | accuracy | No clear effect |
| Add the mixer to correlate the layer latent codes (ablation) | HyperGAN · CIFAR-10 | parameter diversity | Helped |
| Remove both the mixer and the discriminator (ablation) | HyperGAN · CIFAR-10 | parameter diversity | Hurt |
| Use HyperGAN-generated ensembles rather than the evaluated baseline approaches | HyperGAN · notMNIST out-of-distribution detection | predictive entropy | Helped |
| Use HyperGAN-generated ensembles rather than standard ensembles | HyperGAN · adversarial-example detection | predictive entropy | Helped |
Closest papers by meaning
- RareGAN: Generating Samples for Rare Classes (AAAI 2022 · similarity 0.89)
- AdaGAN: Boosting Generative Models (NeurIPS 2017 · similarity 0.89)
- A Unified View of cGANs with and without Classifiers (NeurIPS 2021 · similarity 0.88)
- A Large-Scale Study on Regularization and Normalization in GANs (ICML 2019 · similarity 0.88)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.88)
- Rebooting ACGAN: Auxiliary Classifier GANs with Stable Training (NeurIPS 2021 · similarity 0.87)
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.