AI Paper Atlas

Research map / Benchmarks / CIFAR-100, ResNet-20

CIFAR-100, ResNet-20 leaderboard

11 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on CIFAR-100, ResNet-20. Best accuracy: 77.92 (Hybrid searched policies, CVPR 2021).

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-100, ResNet-20 · accuracy

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

#accuracy ↑MethodPaperTopic
177.92Hybrid searched policies
ResNet20
Privacy-Preserving Collaborative Learning With Automatic Transformation Search
CVPR 2021 · Table: PSNR and model accuracy
Privacy attacks and secure learning
277.37DEO★-mSGD×P10
ResNet20
Non-reversible Parallel Tempering for Deep Posterior Approximation
AAAI 2023 · Table 1
Distributed and compressed
372.22BOSS
ResNet-20
Bayesian Optimization Meets Self-Distillation
ICCV 2023 · Table 2
Knowledge distillation
469.82Our Work-5-T
ResNet-20
Optimal Conversion of Conventional Artificial Neural Networks to Spiking Neural Networks
ICLR 2021 · Table 1
Spiking networks
568.97LookupNet
ResNet-20
Deep Lookup Network
TPAMI 2025 · Table 6
CNN design
668.9TAS
ResNet-20
Network Pruning via Transformable Architecture Search
NeurIPS 2019 · Table: Comparison of different
Neural architecture search
767.86Augmented dropout p=0.1
ResNet-20
Demystifying Dropout
ICML 2019 · Table 5
RNNs and normalization
865.99signSGD-SIM
ResNet20
Momentum Ensures Convergence of SIGNSGD under Weaker Assumptions
ICML 2023 · Table 3
Distributed and compressed
964.95Itr-PSO
ResNet-20
Pruning via Sparsity-indexed ODE: a Continuous Sparsity Viewpoint
ICML 2023 · Table 1
Pruning and efficient tuning
1054.7With dropoutImplicit Regularization of Dropout
TPAMI 2024 · Table 1
Implicit bias and generalization
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.