List your company
ResearchFormerly WizWand

Image Classification on SST2 Rendered (top-1 accuracy)

68.4Top-1 AccuracyInternVL-C
47.91253.23158.5563.869Feb 6, 2024Jun 11, 2024Oct 15, 2024Feb 18, 2025Jun 24, 2025Oct 28, 2025Mar 3, 2026
Updated 8mo ago

Evaluation results

MethodLinks
InternVL-CZero-shot=true2024.0268.4
EVA-CLIP-18BZero-shot=true2024.0267.5
EVA-CLIP-8BZero-shot=true2024.0266.2
OpenCLIP-G/14Zero-shot=true2024.0265.3
EVA-02-CLIP-E/14+Zero-shot=true2024.0262.6
CLIPImage Encoder=ViT-B/16, Zero-shot=true2026.0260.5
OpenCLIPImage Encoder=ViT-B/16, Zero-shot=true2026.0259.9
EVA-01-CLIP-g/14+Zero-shot=true2024.0258.6
EVA-01-CLIP-g/14Zero-shot=true2024.0258.4
Standard CLIPEvaluation Protocol=zero-shot2024.1156
DFN5B-CLIP-H/14+Zero-shot=true2024.0255.5
DFN5B-CLIP-H/14Zero-shot=true2024.0254.8
CLIPZero-shot=true, Backbone=ViT-B/16, Pre-training Dataset=Laion100M, Epochs=302026.0354.6
ITOZero-shot=true, Backbone=ViT-B/16, Pre-training Dataset=Laion100M, Epochs=302026.0354.4
ITOBackbone=ViT-L/16, Pre-training Dataset=DataComp-1B, Training Epochs=1, Evaluation Protocol=Zero-shot2026.0353.7
ITO sub2Pre-training Dataset=DataComp-1B, Backbone=ViT-B/16, Training Epochs=10 epochs, Zero-shot protocol=EVA-CLIP2026.0352.4
ITO sub2Backbone=ViT-L/16, Pre-training Dataset=DataComp-1B, Training Epochs=1, Evaluation Protocol=Zero-shot2026.0352
CLIPBackbone=ViT-L/16, Pre-training Dataset=DataComp-1B, Training Epochs=1, Evaluation Protocol=Zero-shot2026.0351.8
CLIPPre-training Dataset=DataComp-1B, Backbone=ViT-B/16, Training Epochs=10 epochs, Zero-shot protocol=EVA-CLIP2026.0351.3
SLIPZero-shot=true, Backbone=ViT-B/16, Pre-training Dataset=Laion100M, Epochs=302026.0350.6
ITOPre-training Dataset=DataComp-1B, Backbone=ViT-B/16, Training Epochs=10 epochs, Zero-shot protocol=EVA-CLIP2026.0350.5
ITOPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.0350.2
CLIP-MapsmallImage Encoder=ViT-8M/16, Zero-shot=true2026.0250.1
FLAIRBackbone=ViT-B/16, Pre-training Dataset=CC3M-recap, Evaluation Protocol=Zero-shot2026.0350.1
ITO sub2Backbone=ViT-B/16, Pre-training Dataset=CC3M-recap, Evaluation Protocol=Zero-shot2026.0350.1
ITO sub3Backbone=ViT-B/16, Pre-training Dataset=CC3M-recap, Evaluation Protocol=Zero-shot2026.0350.1
SLIPPre-train dataset=CC3M, Backbone=ViT-B/16, Training Epochs=30, Evaluation Protocol=Zero-shot2026.0350.1
SigLIPPre-train dataset=CC3M, Backbone=ViT-B/16, Training Epochs=30, Evaluation Protocol=Zero-shot2026.0350.1
FLAIRPre-train dataset=CC3M, Backbone=ViT-B/16, Training Epochs=30, Evaluation Protocol=Zero-shot2026.0350.1
ITOPre-train dataset=CC3M, Backbone=ViT-B/16, Training Epochs=30, Evaluation Protocol=Zero-shot2026.0350.1
Text2ConceptEvaluation Protocol=zero-shot2024.1150
B-cosified RN-50 CLIPPre-training Dataset=ImageNet [16], Learning Scheduler=Cosine, Evaluation Protocol=zero-shot2024.1150
B-cosified RN-50 CLIPPre-training Dataset=ImageNet [16], Learning Scheduler=Cyclic, Evaluation Protocol=zero-shot2024.1150
B-cosified RN-50 CLIPPre-training Dataset=CC3M [47], Learning Scheduler=Cosine, Evaluation Protocol=zero-shot2024.1150
B-cosified RN-50 CLIPPre-training Dataset=CC3M [47], Learning Scheduler=Cyclic, Evaluation Protocol=zero-shot2024.1150
CLIP-MaptinyImage Encoder=ViT-0.8M/16, Zero-shot=true2026.0250
†TinyCLIPImage Encoder=ViT-8M/16, Zero-shot=true2026.0250
CLIPPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.0350
ITO sub2Pre-train dataset=CC3M, Backbone=ViT-B/16, Training Epochs=30, Evaluation Protocol=Zero-shot2026.0350
TinyCLIPImage Encoder=ViT-39M/16, Zero-shot=true2026.0249.9
FLAIRPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.0349.9
ITO sub2Pre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.0349.9
CLIPPre-train dataset=CC3M, Backbone=ViT-B/16, Training Epochs=30, Evaluation Protocol=Zero-shot2026.0349.9
SigLIPPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.0349.6
†TinyCLIPImage Encoder=ViT-0.8M/16, Zero-shot=true2026.0249.3
SLIPPre-training Dataset=CC12M, Backbone=ViT-B/16, Epochs=30, Evaluation protocol=Zero-shot2026.0348.8
CLIP-MapbaseImage Encoder=ViT-39M/16, Zero-shot=true2026.0248.7