AI Paper Atlas

Research map / Benchmarks / CIFAR-10 (from scratch)

CIFAR-10 (from scratch) leaderboard

15 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CIFAR-10 (from scratch). Best accuracy: 98.28 (Strip-MLP-B, ICCV 2023).

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 (from scratch) · accuracy

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

#accuracy ↑MethodPaperTopic
198.28Strip-MLP-BStrip-MLP: Efficient Token Interaction for Vision MLP
ICCV 2023 · Supp. Table
Attention and transformers
297.63LE-SAM
PyramidNet-110
Fix the Loss, Not the Radius: Rethinking the Adversarial Perturbation of Sharpness-Aware Minimization
ICML 2026 · Table 1
Implicit bias and generalization
397.19Nonuniformity-guided (Algorithm 1)
ResNet-58-6
Beyond Sharpness: The Role of Nonuniformity in Generalization
AAAI 2026 · Table 1
Implicit bias and generalization
497.15Align-ASAM
WideResNet28x10
Align-SAM: Seeking Flatter Minima for Better Cross-Subset Alignment
ICLR 2026 · Table 2
Implicit bias and generalization
597.13RSAM (Renyi sharpness)
WideResNet-28-10
Rényi Sharpness: A Novel Sharpness that Strongly Correlates with Generalization
ICLR 2026 · Table 1
Implicit bias and generalization
696.1VVT-SVicinity Vision Transformer
TPAMI 2023 · Table 3
Attention and transformers
794.1Mechanic-Momentum
ResNet-18
Mechanic: A Learning Rate Tuner
NeurIPS 2023 · Table 2
Optimizers
892.84IGC-L24M2
depth 20
Interleaved Group Convolutions
ICCV 2017 · Table 4
CNN design
992.81Procedural warm-up (k-Dyck)
ViT-T/16
Can You Learn to See Without Images? Procedural Warm-Up for Vision Transformers
CVPR 2026 · Table 1
Attention and transformers
1091.58LeRaC
Wide-ResNet-50
Learning Rate Curriculum
IJCV 2024 · Table 4
CNN design
5 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.