AI Paper Atlas

Research map / Benchmarks / CIFAR-10 32×32

CIFAR-10 32×32 leaderboard

11 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CIFAR-10 32×32. Best FID: 1.96 (CoSIM, 6 NFE, ICML 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

CIFAR-10 32×32 · FID

11 papers, each with its best reported number for its own method; lower is better.

#FID ↓MethodPaperTopic
11.96CoSIM, 6 NFE
VP-EDM score network
Continuous Semi-Implicit Models
ICML 2025 · Table: Unconditional generatio
Fast diffusion sampling
22.44EDM + Ours
EDM
Denoising Task Difficulty-based Curriculum for Training Diffusion Models
ICLR 2025 · Comparative results table
Fast diffusion sampling
32.47DDPM++ (VP, FID) + ST
DDPM++ (VP)
Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation
ICML 2022 · Table: Performance comparisons
Fast diffusion sampling
42.55ADM w/ SMD
ADM
Soft Mixture Denoising: Beyond the Expressive Bottleneck of Diffusion Models
ICLR 2024 · Table: SMD consistently improv
Fast diffusion sampling
53.19Diffusion StyleGAN2
StyleGAN2
Diffusion-GAN: Training GANs with Diffusion
ICLR 2023 · Table 1
GANs
64.01Wavelet Diffusion
DDGAN
Wavelet Diffusion Models Are Fast and Scalable Image Generators
CVPR 2023 · CIFAR-10 results table
Fast diffusion sampling
74.35Ours - N=2
DDPM
DeepCache: Accelerating Diffusion Models for Free
CVPR 2024 · Table 2
Fast diffusion sampling
84.76RFMReflected Flow Matching
ICML 2024 · Table 3
Variational and generative sampling
95.47FlowDCN-S/2Exploring DCN-like architecture for fast image generation with arbitrary resolution
NeurIPS 2024 · Table: Comparisons with SiT
Autoregressive and latent generation
105.71EB-Diff-Train (EB), 30% magnitude pruning
DDPM
Early-Bird Diffusion: Investigating and Leveraging Timestep-Aware Early-Bird Tickets in Diffusion Models for Efficient Training
CVPR 2025 · Table 4
Fast diffusion sampling
1 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.