AI Paper Atlas

Research map / Benchmarks / CIFAR-10 classification

CIFAR-10 classification leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CIFAR-10 classification. Best accuracy: 95.47 (LIFT (10 augmentations), ICCV 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 classification · accuracy

10 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
195.47LIFT (10 augmentations)LIFT: Latent Implicit Functions for Task- and Data-Agnostic Encoding
ICCV 2025 · Table 3
Autoregressive and latent generation
295.34Cosine-based softmax loss
ResNet34
Simple and Effective Out-of-Distribution Detection via Cosine-based Softmax Loss
ICCV 2023 · Table 4
OOD and anomaly detection
394.5SPM-C (100%)Designing to Forget: Deep Semi-parametric Models for Unlearning
CVPR 2026 · Table 1
Recognition and long-tail
494.2FlowCon as classifier
ResNet18
FlowCon: Out-of-Distribution Detection using Flow-based Contrastive Learning
ECCV 2024 · Table 4
OOD and anomaly detection
593.02Stego
ResNet18
Steganography of Steganographic Networks
AAAI 2023 · Table: The fidelity of the pro
RNNs and normalization
690.6SymmFlow (25 steps)Symmetrical Flow Matching: Unified Image Generation, Segmentation, and Classification with Score-Based Generative Models
AAAI 2026 · Table 3
Autoregressive and latent generation
789.2I-EDL
VGG16
Uncertainty Estimation by Fisher Information-based Evidential Deep Learning
ICML 2023 · CIFAR10 misclassification tabl
Recognition and long-tail
888.07Original + PNL (pseudo-non-linear)
ResNet-18
Pseudo-Non-Linear Data Augmentation: A Constrained Energy Minimization Viewpoint
ICLR 2026 · Table 1
Variational and generative sampling
986.7IRDM N=16
single ODE block
Interpolation Technique to Speed Up Gradients Propagation in Neural ODEs
NeurIPS 2020 · text
Variational and generative sampling
1071.97Proposed
CNN1
Heterogeneous Knowledge Distillation Using Information Flow Modeling
CVPR 2020 · Table: Classification Evaluati
Knowledge distillation

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