List your company

Image Classification on STL-10 (Top-1 Accuracy)

99.6Top-1 AccuracyEVA-CLIP-18B
84.873688.696892.5296.3432Oct 14, 2021Jul 9, 2022Apr 4, 2023Dec 28, 2023Sep 22, 2024Jun 17, 2025Mar 13, 2026

Evaluation results

MethodLinks
EVA-CLIP-18BZero-shot=true2024.0299.6—
EVA-01-CLIP-g/14+Zero-shot=true2024.0299.5—
InternVL-CZero-shot=true2024.0299.4—
EVA-02-CLIP-E/14+Zero-shot=true2024.0299.3—
EVA-CLIP-8BZero-shot=true2024.0299.3—
DFN5B-CLIP-H/14+Zero-shot=true2024.0299.1—
EVA-01-CLIP-g/14Zero-shot=true2024.0298.9—
DFN5B-CLIP-H/14Zero-shot=true2024.0298.9—
Mimetic init.Backbone=ViT-B, Pre-training=ImageNet-1K, Initialization/Warm-up=Mimetic init. [51], Evaluation Protocol=Fine-tuned2025.1198.81—
Procedural warm-upBackbone=ViT-B, Pre-training=ImageNet-1K, Initialization/Warm-up=Procedural warm-up (ours), Evaluation Protocol=Fine-tuned2025.1198.66—
Random init.Backbone=ViT-B, Pre-training=ImageNet-1K, Initialization/Warm-up=Random init. [10], Evaluation Protocol=Fine-tuned2025.1198.55—
OpenCLIP-G/14Zero-shot=true2024.0298.5—
FractalDBBackbone=ViT-B, Pre-training=ImageNet-1K, Initialization/Warm-up=FractalDB [23] warm-up, Evaluation Protocol=Fine-tuned2025.1198.46—
xCLIPLinear probing=true, Backbone=ViT-B/16, Pre-train=IT35M2022.1098.1—
OpenCLIPImage Encoder=ViT-B/16, Zero-shot=true2026.0297.9—
CLIPBackbone=CLIP ViT-L-14, zero-shot=true2023.0597.7—
POS PCABackbone=CLIP ViT-L-14, zero-shot=true2023.0597.7—
POS PGABackbone=CLIP ViT-L-14, zero-shot=true2023.0597.7—
CLIPImage Encoder=ViT-B/16, Zero-shot=true2026.0297.6—
nCLIPLinear probing=true, Backbone=ViT-B/16, Pre-train=IT35M2022.1097.4—
TinyCLIPImage Encoder=ViT-39M/16, Zero-shot=true2026.0297.3—
CLIP-MapbaseImage Encoder=ViT-39M/16, Zero-shot=true2026.0297.3—
CLIPLinear probing=true, Backbone=ViT-B/16, Pre-train=IT35M2022.1097.1—
Noun SubmanifoldBackbone=CLIP ViT-L-14, zero-shot=true2023.0596.9—
Noun SubmanifoldBackbone=CLIP ViT-B-32, Evaluation Protocol=Zero-shot2023.0596.3—
CLIPBackbone=CLIP ViT-B-32, Evaluation Protocol=Zero-shot2023.0596.1—
POS PCABackbone=CLIP ViT-B-32, Evaluation Protocol=Zero-shot2023.0596.1—
POS PGABackbone=CLIP ViT-B-32, Evaluation Protocol=Zero-shot2023.0595.8—
IL-CLIPPre-training Dataset=CC12M, Backbone=ViT-B/32, Zero-shot=true2024.0494.1—
†TinyCLIPImage Encoder=ViT-8M/16, Zero-shot=true2026.0292.5—
SAMix-CBackbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=SAMix-C, Predefined Cluster Information=true2021.1192.45—
SAMix-IBackbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=SAMix-I2021.1192.41—
CLIP-MapsmallImage Encoder=ViT-8M/16, Zero-shot=true2026.0292.3—
MoCo.V2(C) + Inter-Intra*Backbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=Inter-Intra*, Predefined Cluster Information=true, Modified by Authors=true2021.1192.24—
MoCo.V2 + MixupBackbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=Mixup2021.1192.2—
B-12/Wi-1024 + DA + TTABlocks=12, Width=1024, Data Augmentation=true, Test-Time Augmentation=true, Pre-training=ImageNet21k, Evaluation Protocol=Fine-tuning2023.0692.2—
BYOLBackbone=ResNet-18, Epochs=1000, linear=true2023.1291.99—
MoCo.V2(C) + PuzzleMix*Backbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=PuzzleMix*, Predefined Cluster Information=true, Modified by Authors=true2021.1191.98—
Ours (IE-CL)Backbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0391.9—
MoCo.V2 + ManifoldMixBackbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=ManifoldMix2021.1191.65—
MoCo.V2 + FMix*Backbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=FMix*, Modified by Authors=true2021.1191.56—
MinEntBackbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0391.5—
MoCo-v3Backbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0391.4—
S3OCBackbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0391.4—
OURS_GCProtocol=Linear probing2023.0391.31—
BYOLBackbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0391.3—
Mixed Barlow TwinsBackbone=ResNet-18, Epochs=1000, linear=true2023.1291.04—
Mixed Barlow TwinsBackbone=ResNet-18, Epochs=2000, linear=true2023.1291.02—
OURS_SepProtocol=Linear probing2023.0390.96—
MoCo.V2 + ResizeMix*Backbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=ResizeMix*, Modified by Authors=true2021.1190.83—
MoCo.V2 + CutMixBackbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=CutMix2021.1190.81—
MoCo.V2 + SaliencyMixBackbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=SaliencyMix2021.1190.77—
OURS_BrProtocol=Linear probing2023.0390.69—
SimCLRBackbone=ResNet-18, Epochs=1000, linear=true2023.1290.51—
SimSiamBackbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0390.5—
W-MSEBackbone=ResNet-18, Epochs=1000, linear=true2023.1290.36—
Codebook-CLIPPre-training Dataset=CC12M, Backbone=ViT-B/32, Zero-shot=true2024.0490.3—
SimCLRBackbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0390.1—
B-12/Wi-1024 + DABlocks=12, Width=1024, Data Augmentation=true, Test-Time Augmentation=false, Pre-training=ImageNet21k, Evaluation Protocol=Fine-tuning2023.0689.9—
NegCLIPPre-training Dataset=CC12M, Backbone=ViT-B/32, Zero-shot=true2024.0489.7—
MoCo.V2Backbone=ResNet-50, Pre-training Epochs=800, SSL Method=MoCo.V22021.1189.68—
OursDistortion=Random Missing Pixels (RM), Evaluation Protocol=Linear Probing2021.1089.44—
CLIPPre-training Dataset=CC12M, Backbone=ViT-B/32, Zero-shot=true2024.0489.3—
Mixed Barlow TwinsBackbone=ResNet-18, Epochs=2000, k-NN=true2023.1288.94—
SAMix-CBackbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=SAMix-C, Predefined Cluster Information=true2021.1188.91—
MoCo-v2Backbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0388.9—
SAMix-IBackbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=SAMix-I2021.1188.87—
SWAV (C)Backbone=ResNet-50, Pre-training Epochs=800, SSL Method=SWAV, Predefined Cluster Information=true2021.1188.79—
BYOLBackbone=ResNet-18, Epochs=1000, k-NN=true2023.1288.64—
SAMix-CBackbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=SAMix-C, Predefined Cluster Information=true2021.1188.63—
SAMix-IBackbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=SAMix-I2021.1188.58—
MoCo.V2(C) + PuzzleMix*Backbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=PuzzleMix*, Predefined Cluster Information=true, Modified by Authors=true2021.1188.4—
MoCo.V2(C) + Inter-Intra*Backbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=Inter-Intra*, Predefined Cluster Information=true, Modified by Authors=true2021.1188.33—
MoCo.V2 + MixupBackbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=Mixup2021.1188.24—
Mixed Barlow TwinsBackbone=ResNet-18, Epochs=1000, k-NN=true2023.1288.23—
Barlow TwinsBackbone=ResNet-18, Epochs=1000, linear=true2023.1288.21—
MoCo.V2 + FMix*Backbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=FMix*, Modified by Authors=true2021.1188.14—
MoCo.V2(C) + PuzzleMix*Backbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=PuzzleMix*, Predefined Cluster Information=true, Modified by Authors=true2021.1188.07—
MoCo.V2 + ManifoldMixBackbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=ManifoldMix2021.1188.06—
MoCo.V2 + MixupBackbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=Mixup2021.1187.93—
MoCo.V2(C) + Inter-Intra*Backbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=Inter-Intra*, Predefined Cluster Information=true, Modified by Authors=true2021.1187.85—
MoCo.V2 + ManifoldMixBackbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=ManifoldMix2021.1187.7—
IL-CLIPPre-training Dataset=CC3M, Backbone=ViT-B/32, Zero-shot=true2024.0487.7—
W-MSEBackbone=ResNet-18, Batch Size=256, Evaluation Protocol=Linear probe2026.0387.7—
MoCo.V2 + FMix*Backbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=FMix*, Modified by Authors=true2021.1187.68—
CyCLIPProtocol=Linear probing2023.0387.61—
MoCo.V2 + CutMixBackbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=CutMix2021.1187.6—
MoCo.V2 + CutMixBackbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=CutMix2021.1187.51—
MoCo.V2 + SaliencyMixBackbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=SaliencyMix2021.1187.35—
MoCo.V2 + SaliencyMixBackbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=SaliencyMix2021.1187.27—
MoCo.V2 + ResizeMix*Backbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V2, Mixup Variant=ResizeMix*, Modified by Authors=true2021.1187.25—
CLIPProtocol=Linear probing2023.0387.22—
W-MSEBackbone=ResNet-18, Epochs=1000, k-NN=true2023.1287.1—
MoCo.V2 + ResizeMix*Backbone=ResNet-50, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=ResizeMix*, Modified by Authors=true2021.1186.88—
OursDistortion=Additive Gaussian Noise (GN), Evaluation Protocol=Linear Probing2021.1086.3—
SimCLRBackbone=ResNet-18, Epochs=1000, k-NN=true2023.1285.68—
MoCo.V2Backbone=ResNet-18, Pre-training Epochs=800, SSL Method=MoCo.V22021.1185.64—
SAMix-CBackbone=ResNet-18, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=SAMix-C, Predefined Cluster Information=true2021.1185.6—
SWAV (C)Backbone=ResNet-18, Pre-training Epochs=800, SSL Method=SWAV, Predefined Cluster Information=true2021.1185.56—
SAMix-IBackbone=ResNet-18, Pre-training Epochs=400, SSL Method=MoCo.V2, Mixup Variant=SAMix-I2021.1185.44—