{"url":"/sota/image-classification-on-imagenet-real","task":{"name":"Image Classification","url":"/task/image-classification","note":null},"dataset":{"name":"ImageNet ReaL","url":null},"category":"Computer Vision","categories":["Adversarial","Computer Vision"],"category_note":null,"description":"**Image Classification** is a fundamental task in vision recognition that aims to understand and categorize an image as a whole under a specific label. Unlike [object detection](/task/object-detection), which involves classification and location of multiple objects within an image, image classification typically pertains to single-object images. When the classification becomes highly detailed or reaches instance-level, it is often referred to as [image retrieval](/task/image-retrieval), which also involves finding similar images in a large database.\r\n\r\n\r\n<span class=\"description-source\">Source: [Metamorphic Testing for Object Detection Systems ](https://arxiv.org/abs/1912.12162)</span>","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["Accuracy","Params","Top 1 Accuracy","Number of params"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Accuracy":"higher","Params":"lower","Top 1 Accuracy":"higher","Number of params":"lower"}},"counts":{"rows":57,"rows_with_code":56,"rows_with_paper_page":57,"rows_dated":57,"rows_using_additional_data":21},"rows":[{"rank_in_archive_order":1,"model":"Baseline (ViT-G/14)","metrics":{"Accuracy":"91.78%"},"uses_additional_data":true,"paper_date":"2022-03-10","paper":"/paper/model-soups-averaging-weights-of-multiple","paper_url":"https://arxiv.org/abs/2203.05482v3","paper_title":"Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time","code":"https://github.com/mlfoundations/model-soups","n_code_links":6,"syntology":{"n_ran":5,"n_unverified":12,"n_samples":17,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"ViTAE-H\n(MAE, 512)","metrics":{"Accuracy":"91.2%","Params":"644M"},"uses_additional_data":false,"paper_date":"2022-02-21","paper":"/paper/vitaev2-vision-transformer-advanced-by","paper_url":"https://arxiv.org/abs/2202.10108v2","paper_title":"ViTAEv2: Vision Transformer Advanced by Exploring Inductive Bias for Image Recognition and Beyond","code":"https://github.com/ViTAE-Transformer/ViTDet","n_code_links":8,"syntology":{"n_ran":15,"n_unverified":10,"n_samples":25,"n_pointer_only_licence":2}},{"rank_in_archive_order":3,"model":"Model soups (ViT-G/14)","metrics":{"Accuracy":"91.20%","Params":"1843M"},"uses_additional_data":true,"paper_date":"2022-03-10","paper":"/paper/model-soups-averaging-weights-of-multiple","paper_url":"https://arxiv.org/abs/2203.05482v3","paper_title":"Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time","code":"https://github.com/mlfoundations/model-soups","n_code_links":6,"syntology":{"n_ran":5,"n_unverified":12,"n_samples":17,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"Meta Pseudo Labels (EfficientNet-B6-Wide)","metrics":{"Accuracy":"91.12%"},"uses_additional_data":true,"paper_date":"2020-03-23","paper":"/paper/meta-pseudo-labels","paper_url":"https://arxiv.org/abs/2003.10580v4","paper_title":"Meta Pseudo Labels","code":"https://github.com/google-research/google-research/tree/master/meta_pseudo_labels","n_code_links":9,"syntology":{"n_ran":5,"n_unverified":9,"n_samples":14,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"MAWS (ViT-6.5B)","metrics":{"Accuracy":"91.1%"},"uses_additional_data":true,"paper_date":"2023-03-23","paper":"/paper/the-effectiveness-of-mae-pre-pretraining-for","paper_url":"https://arxiv.org/abs/2303.13496v3","paper_title":"The effectiveness of MAE pre-pretraining for billion-scale pretraining","code":"https://github.com/facebookresearch/maws","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"TokenLearner L/8 (24+11)","metrics":{"Accuracy":"91.05%","Params":"460M"},"uses_additional_data":true,"paper_date":"2021-06-21","paper":"/paper/tokenlearner-what-can-8-learned-tokens-do-for","paper_url":"https://arxiv.org/abs/2106.11297v4","paper_title":"TokenLearner: What Can 8 Learned Tokens Do for Images and Videos?","code":"https://github.com/google-research/scenic","n_code_links":11,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"Model soups (BASIC-L)","metrics":{"Accuracy":"91.03%","Params":"2440M"},"uses_additional_data":true,"paper_date":"2022-03-10","paper":"/paper/model-soups-averaging-weights-of-multiple","paper_url":"https://arxiv.org/abs/2203.05482v3","paper_title":"Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time","code":"https://github.com/mlfoundations/model-soups","n_code_links":6,"syntology":{"n_ran":5,"n_unverified":12,"n_samples":17,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"Meta Pseudo Labels (EfficientNet-L2)","metrics":{"Accuracy":"91.02%"},"uses_additional_data":true,"paper_date":"2020-03-23","paper":"/paper/meta-pseudo-labels","paper_url":"https://arxiv.org/abs/2003.10580v4","paper_title":"Meta Pseudo Labels","code":"https://github.com/google-research/google-research/tree/master/meta_pseudo_labels","n_code_links":9,"syntology":{"n_ran":5,"n_unverified":9,"n_samples":14,"n_pointer_only_licence":0}},{"rank_in_archive_order":9,"model":"MAWS (ViT-2B)","metrics":{"Accuracy":"90.9%"},"uses_additional_data":true,"paper_date":"2023-03-23","paper":"/paper/the-effectiveness-of-mae-pre-pretraining-for","paper_url":"https://arxiv.org/abs/2303.13496v3","paper_title":"The effectiveness of MAE pre-pretraining for billion-scale pretraining","code":"https://github.com/facebookresearch/maws","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"FixEfficientNet-L2","metrics":{"Accuracy":"90.9%","Params":"480M"},"uses_additional_data":true,"paper_date":"2020-03-18","paper":"/paper/fixing-the-train-test-resolution-discrepancy-2","paper_url":"https://arxiv.org/abs/2003.08237v5","paper_title":"Fixing the train-test resolution discrepancy: FixEfficientNet","code":"https://github.com/facebookresearch/FixRes","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"ViT-G/14","metrics":{"Accuracy":"90.81%"},"uses_additional_data":true,"paper_date":"2021-06-08","paper":"/paper/scaling-vision-transformers","paper_url":"https://arxiv.org/abs/2106.04560v2","paper_title":"Scaling Vision Transformers","code":"https://github.com/google-research/big_vision","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"MAWS (ViT-H)","metrics":{"Accuracy":"90.8%"},"uses_additional_data":true,"paper_date":"2023-03-23","paper":"/paper/the-effectiveness-of-mae-pre-pretraining-for","paper_url":"https://arxiv.org/abs/2303.13496v3","paper_title":"The effectiveness of MAE pre-pretraining for billion-scale pretraining","code":"https://github.com/facebookresearch/maws","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"SWAG (RegNetY 128GF)","metrics":{"Accuracy":"90.7%"},"uses_additional_data":true,"paper_date":"2022-01-20","paper":"/paper/revisiting-weakly-supervised-pre-training-of","paper_url":"https://arxiv.org/abs/2201.08371v2","paper_title":"Revisiting Weakly Supervised Pre-Training of Visual Perception Models","code":"https://github.com/facebookresearch/SWAG","n_code_links":2,"syntology":null},{"rank_in_archive_order":14,"model":"CvT-W24 (384 res, ImageNet-22k pretrain)","metrics":{"Accuracy":"90.6%","Number of params":"277M","Top 1 Accuracy":"87.7%"},"uses_additional_data":true,"paper_date":"2021-03-29","paper":"/paper/cvt-introducing-convolutions-to-vision","paper_url":"https://arxiv.org/abs/2103.15808v1","paper_title":"CvT: Introducing Convolutions to Vision Transformers","code":"https://github.com/huggingface/transformers","n_code_links":16,"syntology":{"n_ran":29,"n_unverified":18,"n_samples":47,"n_pointer_only_licence":4}},{"rank_in_archive_order":15,"model":"VOLO-D5","metrics":{"Accuracy":"90.6%"},"uses_additional_data":false,"paper_date":"2021-06-24","paper":"/paper/volo-vision-outlooker-for-visual-recognition","paper_url":"https://arxiv.org/abs/2106.13112v2","paper_title":"VOLO: Vision Outlooker for Visual Recognition","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":7,"syntology":{"n_ran":1,"n_unverified":5,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":16,"model":"EfficientNet-L2","metrics":{"Accuracy":"90.55%","Params":"480M"},"uses_additional_data":true,"paper_date":"2019-11-11","paper":"/paper/self-training-with-noisy-student-improves","paper_url":"https://arxiv.org/abs/1911.04252v4","paper_title":"Self-training with Noisy Student improves ImageNet classification","code":"https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet","n_code_links":13,"syntology":{"n_ran":5,"n_unverified":19,"n_samples":24,"n_pointer_only_licence":0}},{"rank_in_archive_order":17,"model":"BiT-L","metrics":{"Accuracy":"90.54%","Params":"928M"},"uses_additional_data":true,"paper_date":"2019-12-24","paper":"/paper/large-scale-learning-of-general-visual","paper_url":"https://arxiv.org/abs/1912.11370v3","paper_title":"Big Transfer (BiT): General Visual Representation Learning","code":"https://github.com/google-research/big_transfer","n_code_links":9,"syntology":{"n_ran":3,"n_unverified":7,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":18,"model":"VOLO-D4","metrics":{"Accuracy":"90.5%"},"uses_additional_data":false,"paper_date":"2021-06-24","paper":"/paper/volo-vision-outlooker-for-visual-recognition","paper_url":"https://arxiv.org/abs/2106.13112v2","paper_title":"VOLO: Vision Outlooker for Visual Recognition","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":7,"syntology":{"n_ran":1,"n_unverified":5,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":19,"model":"CAIT-M36-448","metrics":{"Accuracy":"90.2%"},"uses_additional_data":false,"paper_date":"2021-03-31","paper":"/paper/going-deeper-with-image-transformers","paper_url":"https://arxiv.org/abs/2103.17239v2","paper_title":"Going deeper with Image Transformers","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":21,"syntology":{"n_ran":5,"n_unverified":6,"n_samples":11,"n_pointer_only_licence":2}},{"rank_in_archive_order":20,"model":"Mixer-H/14- 448 (JFT-300M pre-train)","metrics":{"Accuracy":"90.18%","Params":"409M"},"uses_additional_data":true,"paper_date":"2021-05-04","paper":"/paper/mlp-mixer-an-all-mlp-architecture-for-vision","paper_url":"https://arxiv.org/abs/2105.01601v4","paper_title":"MLP-Mixer: An all-MLP Architecture for Vision","code":"https://github.com/labmlai/annotated_deep_learning_paper_implementations","n_code_links":49,"syntology":{"n_ran":106,"n_unverified":28,"n_samples":134,"n_pointer_only_licence":36}},{"rank_in_archive_order":21,"model":"FixEfficientNet-B8","metrics":{"Accuracy":"90.0%","Params":"87M"},"uses_additional_data":false,"paper_date":"2020-03-18","paper":"/paper/fixing-the-train-test-resolution-discrepancy-2","paper_url":"https://arxiv.org/abs/2003.08237v5","paper_title":"Fixing the train-test resolution discrepancy: FixEfficientNet","code":"https://github.com/facebookresearch/FixRes","n_code_links":1,"syntology":null},{"rank_in_archive_order":22,"model":"SEER (RegNet10B)","metrics":{"Accuracy":"89.8%","Params":"10000M"},"uses_additional_data":true,"paper_date":"2022-02-16","paper":"/paper/vision-models-are-more-robust-and-fair-when","paper_url":"https://arxiv.org/abs/2202.08360v2","paper_title":"Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervision","code":"https://github.com/facebookresearch/vissl","n_code_links":1,"syntology":null},{"rank_in_archive_order":23,"model":"FixResNeXt-101 32x48d","metrics":{"Accuracy":"89.73%","Params":"829M"},"uses_additional_data":true,"paper_date":"2019-06-14","paper":"/paper/fixing-the-train-test-resolution-discrepancy","paper_url":"https://arxiv.org/abs/1906.06423v4","paper_title":"Fixing the train-test resolution discrepancy","code":"https://github.com/facebookresearch/FixRes","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":24,"model":"DeiT-B-384","metrics":{"Accuracy":"89.3%","Params":"86M"},"uses_additional_data":false,"paper_date":"2020-12-23","paper":"/paper/training-data-efficient-image-transformers","paper_url":"https://arxiv.org/abs/2012.12877v2","paper_title":"Training data-efficient image transformers & distillation through attention","code":"https://github.com/huggingface/transformers","n_code_links":40,"syntology":{"n_ran":12,"n_unverified":7,"n_samples":19,"n_pointer_only_licence":3}},{"rank_in_archive_order":25,"model":"BiT-M","metrics":{"Accuracy":"89.02%"},"uses_additional_data":true,"paper_date":"2019-12-24","paper":"/paper/large-scale-learning-of-general-visual","paper_url":"https://arxiv.org/abs/1912.11370v3","paper_title":"Big Transfer (BiT): General Visual Representation Learning","code":"https://github.com/google-research/big_transfer","n_code_links":9,"syntology":{"n_ran":3,"n_unverified":7,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":26,"model":"DeiT-B","metrics":{"Accuracy":"88.7%","Params":"86M"},"uses_additional_data":false,"paper_date":"2020-12-23","paper":"/paper/training-data-efficient-image-transformers","paper_url":"https://arxiv.org/abs/2012.12877v2","paper_title":"Training data-efficient image transformers & distillation through attention","code":"https://github.com/huggingface/transformers","n_code_links":40,"syntology":{"n_ran":12,"n_unverified":7,"n_samples":19,"n_pointer_only_licence":3}},{"rank_in_archive_order":27,"model":"Assemble-ResNet152","metrics":{"Accuracy":"88.65%"},"uses_additional_data":false,"paper_date":"2020-01-17","paper":"/paper/compounding-the-performance-improvements-of","paper_url":"https://arxiv.org/abs/2001.06268v2","paper_title":"Compounding the Performance Improvements of Assembled Techniques in a Convolutional Neural Network","code":"https://github.com/clovaai/assembled-cnn","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":10,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":28,"model":"CeiT-S (384 finetune res)","metrics":{"Accuracy":"88.1%"},"uses_additional_data":false,"paper_date":"2021-03-22","paper":"/paper/incorporating-convolution-designs-into-visual","paper_url":"https://arxiv.org/abs/2103.11816v2","paper_title":"Incorporating Convolution Designs into Visual Transformers","code":"https://github.com/rishikksh20/CeiT-pytorch","n_code_links":3,"syntology":{"n_ran":9,"n_unverified":2,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":29,"model":"Sequencer2D-L","metrics":{"Accuracy":"87.9"},"uses_additional_data":false,"paper_date":"2022-05-04","paper":"/paper/sequencer-deep-lstm-for-image-classification","paper_url":"https://arxiv.org/abs/2205.01972v4","paper_title":"Sequencer: Deep LSTM for Image Classification","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":5,"syntology":{"n_ran":4,"n_unverified":5,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":30,"model":"Mixer-H/14 (JFT-300M pre-train)","metrics":{"Accuracy":"87.86%","Params":"409M"},"uses_additional_data":true,"paper_date":"2021-05-04","paper":"/paper/mlp-mixer-an-all-mlp-architecture-for-vision","paper_url":"https://arxiv.org/abs/2105.01601v4","paper_title":"MLP-Mixer: An all-MLP Architecture for Vision","code":"https://github.com/labmlai/annotated_deep_learning_paper_implementations","n_code_links":49,"syntology":{"n_ran":106,"n_unverified":28,"n_samples":134,"n_pointer_only_licence":36}},{"rank_in_archive_order":31,"model":"Assemble ResNet-50","metrics":{"Accuracy":"87.82%"},"uses_additional_data":false,"paper_date":"2020-01-17","paper":"/paper/compounding-the-performance-improvements-of","paper_url":"https://arxiv.org/abs/2001.06268v2","paper_title":"Compounding the Performance Improvements of Assembled Techniques in a Convolutional Neural Network","code":"https://github.com/clovaai/assembled-cnn","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":10,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":32,"model":"NASNet-A Large","metrics":{"Accuracy":"87.56%"},"uses_additional_data":false,"paper_date":"2017-07-21","paper":"/paper/learning-transferable-architectures-for","paper_url":"http://arxiv.org/abs/1707.07012v4","paper_title":"Learning Transferable Architectures for Scalable Image Recognition","code":"https://github.com/tensorflow/models/tree/master/research/slim","n_code_links":17,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":33,"model":"LeViT-384","metrics":{"Accuracy":"87.5%"},"uses_additional_data":false,"paper_date":"2021-04-02","paper":"/paper/levit-a-vision-transformer-in-convnet-s","paper_url":"https://arxiv.org/abs/2104.01136v2","paper_title":"LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference","code":"https://github.com/huggingface/transformers","n_code_links":12,"syntology":{"n_ran":23,"n_unverified":7,"n_samples":30,"n_pointer_only_licence":0}},{"rank_in_archive_order":34,"model":"CeiT-S","metrics":{"Accuracy":"87.3%"},"uses_additional_data":false,"paper_date":"2021-03-22","paper":"/paper/incorporating-convolution-designs-into-visual","paper_url":"https://arxiv.org/abs/2103.11816v2","paper_title":"Incorporating Convolution Designs into Visual Transformers","code":"https://github.com/rishikksh20/CeiT-pytorch","n_code_links":3,"syntology":{"n_ran":9,"n_unverified":2,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":35,"model":"LeViT-256","metrics":{"Accuracy":"86.9%"},"uses_additional_data":false,"paper_date":"2021-04-02","paper":"/paper/levit-a-vision-transformer-in-convnet-s","paper_url":"https://arxiv.org/abs/2104.01136v2","paper_title":"LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference","code":"https://github.com/huggingface/transformers","n_code_links":12,"syntology":{"n_ran":23,"n_unverified":7,"n_samples":30,"n_pointer_only_licence":0}},{"rank_in_archive_order":36,"model":"DeiT-S","metrics":{"Accuracy":"86.8%","Params":"22M"},"uses_additional_data":false,"paper_date":"2020-12-23","paper":"/paper/training-data-efficient-image-transformers","paper_url":"https://arxiv.org/abs/2012.12877v2","paper_title":"Training data-efficient image transformers & distillation through attention","code":"https://github.com/huggingface/transformers","n_code_links":40,"syntology":{"n_ran":12,"n_unverified":7,"n_samples":19,"n_pointer_only_licence":3}},{"rank_in_archive_order":37,"model":"ResNet-152x2-SAM","metrics":{"Accuracy":"86.4%"},"uses_additional_data":false,"paper_date":"2021-06-03","paper":"/paper/when-vision-transformers-outperform-resnets","paper_url":"https://arxiv.org/abs/2106.01548v3","paper_title":"When Vision Transformers Outperform ResNets without Pre-training or Strong Data Augmentations","code":"https://github.com/google-research/vision_transformer","n_code_links":2,"syntology":null},{"rank_in_archive_order":38,"model":"LeViT-192","metrics":{"Accuracy":"85.8%"},"uses_additional_data":false,"paper_date":"2021-04-02","paper":"/paper/levit-a-vision-transformer-in-convnet-s","paper_url":"https://arxiv.org/abs/2104.01136v2","paper_title":"LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference","code":"https://github.com/huggingface/transformers","n_code_links":12,"syntology":{"n_ran":23,"n_unverified":7,"n_samples":30,"n_pointer_only_licence":0}},{"rank_in_archive_order":39,"model":"ResNet50 (A1)","metrics":{"Accuracy":"85.7%","Params":"25M"},"uses_additional_data":false,"paper_date":"2021-10-01","paper":"/paper/resnet-strikes-back-an-improved-training","paper_url":"https://arxiv.org/abs/2110.00476v1","paper_title":"ResNet strikes back: An improved training procedure in timm","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":14,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":40,"model":"LeViT-128","metrics":{"Accuracy":"85.6%"},"uses_additional_data":false,"paper_date":"2021-04-02","paper":"/paper/levit-a-vision-transformer-in-convnet-s","paper_url":"https://arxiv.org/abs/2104.01136v2","paper_title":"LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference","code":"https://github.com/huggingface/transformers","n_code_links":12,"syntology":{"n_ran":23,"n_unverified":7,"n_samples":30,"n_pointer_only_licence":0}},{"rank_in_archive_order":41,"model":"ResMLP-36","metrics":{"Accuracy":"85.6%","Params":"45M"},"uses_additional_data":false,"paper_date":"2021-05-07","paper":"/paper/resmlp-feedforward-networks-for-image","paper_url":"https://arxiv.org/abs/2105.03404v2","paper_title":"ResMLP: Feedforward networks for image classification with data-efficient training","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":19,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":42,"model":"ResMLP-24","metrics":{"Accuracy":"85.3%","Params":"30M"},"uses_additional_data":false,"paper_date":"2021-05-07","paper":"/paper/resmlp-feedforward-networks-for-image","paper_url":"https://arxiv.org/abs/2105.03404v2","paper_title":"ResMLP: Feedforward networks for image classification with data-efficient training","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":19,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":43,"model":"ViT-B/16-SAM","metrics":{"Accuracy":"85.2%"},"uses_additional_data":false,"paper_date":"2021-06-03","paper":"/paper/when-vision-transformers-outperform-resnets","paper_url":"https://arxiv.org/abs/2106.01548v3","paper_title":"When Vision Transformers Outperform ResNets without Pre-training or Strong Data Augmentations","code":"https://github.com/google-research/vision_transformer","n_code_links":2,"syntology":null},{"rank_in_archive_order":44,"model":"ResMLP-12","metrics":{"Accuracy":"84.6%","Params":"15M"},"uses_additional_data":false,"paper_date":"2021-05-07","paper":"/paper/resmlp-feedforward-networks-for-image","paper_url":"https://arxiv.org/abs/2105.03404v2","paper_title":"ResMLP: Feedforward networks for image classification with data-efficient training","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":19,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":45,"model":"Mixer-B/8-SAM","metrics":{"Accuracy":"84.4%"},"uses_additional_data":false,"paper_date":"2021-06-03","paper":"/paper/when-vision-transformers-outperform-resnets","paper_url":"https://arxiv.org/abs/2106.01548v3","paper_title":"When Vision Transformers Outperform ResNets without Pre-training or Strong Data Augmentations","code":"https://github.com/google-research/vision_transformer","n_code_links":2,"syntology":null},{"rank_in_archive_order":46,"model":"kNN-CLIP","metrics":{"Accuracy":"84%"},"uses_additional_data":true,"paper_date":"2022-04-03","paper":"/paper/revisiting-a-knn-based-image-classification","paper_url":"https://arxiv.org/abs/2204.01186v2","paper_title":"Revisiting a kNN-based Image Classification System with High-capacity Storage","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":47,"model":"CeiT-T","metrics":{"Accuracy":"83.6%"},"uses_additional_data":false,"paper_date":"2021-03-22","paper":"/paper/incorporating-convolution-designs-into-visual","paper_url":"https://arxiv.org/abs/2103.11816v2","paper_title":"Incorporating Convolution Designs into Visual Transformers","code":"https://github.com/rishikksh20/CeiT-pytorch","n_code_links":3,"syntology":{"n_ran":9,"n_unverified":2,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":48,"model":"LeViT-128S","metrics":{"Accuracy":"82.6%"},"uses_additional_data":false,"paper_date":"2021-04-02","paper":"/paper/levit-a-vision-transformer-in-convnet-s","paper_url":"https://arxiv.org/abs/2104.01136v2","paper_title":"LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference","code":"https://github.com/huggingface/transformers","n_code_links":12,"syntology":{"n_ran":23,"n_unverified":7,"n_samples":30,"n_pointer_only_licence":0}},{"rank_in_archive_order":49,"model":"DeiT-Ti","metrics":{"Accuracy":"82.1%","Params":"5M"},"uses_additional_data":false,"paper_date":"2020-12-23","paper":"/paper/training-data-efficient-image-transformers","paper_url":"https://arxiv.org/abs/2012.12877v2","paper_title":"Training data-efficient image transformers & distillation through attention","code":"https://github.com/huggingface/transformers","n_code_links":40,"syntology":{"n_ran":12,"n_unverified":7,"n_samples":19,"n_pointer_only_licence":3}},{"rank_in_archive_order":50,"model":"NASNet-A Mobile","metrics":{"Accuracy":"81.15%"},"uses_additional_data":false,"paper_date":"2017-07-21","paper":"/paper/learning-transferable-architectures-for","paper_url":"http://arxiv.org/abs/1707.07012v4","paper_title":"Learning Transferable Architectures for Scalable Image Recognition","code":"https://github.com/tensorflow/models/tree/master/research/slim","n_code_links":17,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":51,"model":"VGG-16 BN","metrics":{"Accuracy":"80.60%"},"uses_additional_data":false,"paper_date":"2014-09-04","paper":"/paper/very-deep-convolutional-networks-for-large","paper_url":"http://arxiv.org/abs/1409.1556v6","paper_title":"Very Deep Convolutional Networks for Large-Scale Image Recognition","code":"https://github.com/tensorflow/models/tree/master/research/slim","n_code_links":305,"syntology":{"n_ran":12,"n_unverified":110,"n_samples":122,"n_pointer_only_licence":4}},{"rank_in_archive_order":52,"model":"VGG-16","metrics":{"Accuracy":"79.01%"},"uses_additional_data":false,"paper_date":"2014-09-04","paper":"/paper/very-deep-convolutional-networks-for-large","paper_url":"http://arxiv.org/abs/1409.1556v6","paper_title":"Very Deep Convolutional Networks for Large-Scale Image Recognition","code":"https://github.com/tensorflow/models/tree/master/research/slim","n_code_links":305,"syntology":{"n_ran":12,"n_unverified":110,"n_samples":122,"n_pointer_only_licence":4}},{"rank_in_archive_order":53,"model":"AlexNet","metrics":{"Accuracy":"62.88%"},"uses_additional_data":false,"paper_date":"2012-12-01","paper":"/paper/imagenet-classification-with-deep","paper_url":"http://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks","paper_title":"ImageNet Classification with Deep Convolutional Neural Networks","code":"https://github.com/pytorch/vision","n_code_links":23,"syntology":null},{"rank_in_archive_order":54,"model":"ViT-L @384 (DeiT III, 21k)","metrics":{"Number of params":"304M","Top 1 Accuracy":"87.7%"},"uses_additional_data":false,"paper_date":"2022-04-14","paper":"/paper/deit-iii-revenge-of-the-vit","paper_url":"https://arxiv.org/abs/2204.07118v1","paper_title":"DeiT III: Revenge of the ViT","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":12,"syntology":null},{"rank_in_archive_order":55,"model":"ViT-H @224 (DeiT III, 21k)","metrics":{"Number of params":"632M","Top 1 Accuracy":"87.2%"},"uses_additional_data":false,"paper_date":"2022-04-14","paper":"/paper/deit-iii-revenge-of-the-vit","paper_url":"https://arxiv.org/abs/2204.07118v1","paper_title":"DeiT III: Revenge of the ViT","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":12,"syntology":null},{"rank_in_archive_order":56,"model":"ViT-L @224 (DeiT III, 21k)","metrics":{"Top 1 Accuracy":"87.0%"},"uses_additional_data":false,"paper_date":"2022-04-14","paper":"/paper/deit-iii-revenge-of-the-vit","paper_url":"https://arxiv.org/abs/2204.07118v1","paper_title":"DeiT III: Revenge of the ViT","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":12,"syntology":null},{"rank_in_archive_order":57,"model":"ResMLP-B24/8 (22k)","metrics":{"Top 1 Accuracy":"84.4%"},"uses_additional_data":false,"paper_date":"2021-05-07","paper":"/paper/resmlp-feedforward-networks-for-image","paper_url":"https://arxiv.org/abs/2105.03404v2","paper_title":"ResMLP: Feedforward networks for image classification with data-efficient training","code":"https://github.com/rwightman/pytorch-image-models","n_code_links":19,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0}}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,795 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6795,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2785},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":41,"rows_with_any_sample_ran":37,"distinct_papers_with_graph_line":20,"distinct_papers_with_any_sample_ran":17,"samples_over_distinct_papers":{"n_ran":240,"n_unverified":266,"n_samples":506,"n_pointer_only_licence":54,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":530,"n_unverified":528,"n_samples":1058,"n_pointer_only_licence":104,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}