# Image Classification on ImageNet V2 1.0 (val) Source: [Image Classification on ImageNet V2 1.0 (val)]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:06:58.524Z Evaluation of Image Classification performance on the validation split of the ImageNet V2 dataset (version 1.0), measured using Top-1 Accuracy. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Image Classification | | Dataset (including any stated split) | ImageNet V2 1.0 (val) | | Metric | Top-1 Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 29 of 29 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Top-1 Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Top-1 Accuracy | Result date | Report paper | | --- | --- | --- | --- | | ViT-L↑384 (DeiT III) | 76.7 | Not recorded | 68fa6c591e209131cea3378d | | EfficientNetV2-L | 76.3 | Not recorded | 68fa6c591e209131cea3378d | | ViT-H (DeiT III) | 75.9 | Not recorded | 68fa6c591e209131cea3378d | | EfficientNetV2-M | 75.5 | Not recorded | 68fa6c591e209131cea3378d | | ConvNeXt-L-384 | 75.3 | Not recorded | 68fa6c591e209131cea3378d | | ViT-L (DeiT III) | 75.1 | Not recorded | 68fa6c591e209131cea3378d | | ViT-B↑384 (DeiT III) | 74.8 | Not recorded | 68fa6c591e209131cea3378d | | ConvNeXt-B-384 | 74.7 | Not recorded | 68fa6c591e209131cea3378d | | EfficientNetV2-S | 74 | Not recorded | 68fa6c591e209131cea3378d | | ConvNeXt-L-224 | 74 | Not recorded | 68fa6c591e209131cea3378d | | PatchConvNet-B120-224 | 73.9 | Not recorded | 68fa6c591e209131cea3378d | | EfficientNet-B5 | 73.6 | Not recorded | 68fa6c591e209131cea3378d | | ViT-B (DeiT III) | 73.6 | Not recorded | 68fa6c591e209131cea3378d | | ConvNeXt-B-224 | 73.4 | Not recorded | 68fa6c591e209131cea3378d | | ViT-S↑384 (DeiT III) | 73.1 | Not recorded | 68fa6c591e209131cea3378d | | PatchConvNet-B60-224 | 72.6 | Not recorded | 68fa6c591e209131cea3378d | | RegNetY-16GF | 72.4 | Not recorded | 68fa6c591e209131cea3378d | | EfficientNet-B4 | 72.3 | Not recorded | 68fa6c591e209131cea3378d | | Swin-S-224 | 71.8 | Not recorded | 68fa6c591e209131cea3378d | | RegNetY-8GF | 71.1 | Not recorded | 68fa6c591e209131cea3378d | | PatchConvNet-S60-224 | 71 | Not recorded | 68fa6c591e209131cea3378d | | RegNetY-4GF | 70.7 | Not recorded | 68fa6c591e209131cea3378d | | ResNet-152 | 70.6 | Not recorded | 68fa6c591e209131cea3378d | | ViT-S (DeiT III) | 70.5 | Not recorded | 68fa6c591e209131cea3378d | | ResNet-101 | 70.3 | Not recorded | 68fa6c591e209131cea3378d | | Swin-T-224 | 69.5 | Not recorded | 68fa6c591e209131cea3378d | | ResMLP-B24 | 69 | Not recorded | 68fa6c591e209131cea3378d | | ResNet-50 | 68.7 | Not recorded | 68fa6c591e209131cea3378d | | Mixer-B/16 | 63.2 | Not recorded | 68fa6c591e209131cea3378d | ### ViT-L↑384 (DeiT III) Top-1 Accuracy: 76.7. Result ID: 698f07893a6c314849f50a71. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"304.8M","FLOPs":"191.2G","Resolution":"384"}. This record does not assert independent reproduction. ### EfficientNetV2-L Top-1 Accuracy: 76.3. Result ID: 698f07893a6c314849f50a60. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"118.5M","FLOPs":"53.0G"}. This record does not assert independent reproduction. ### ViT-H (DeiT III) Top-1 Accuracy: 75.9. Result ID: 698f07893a6c314849f50a72. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"632.1M","FLOPs":"167.4G"}. This record does not assert independent reproduction. ### EfficientNetV2-M Top-1 Accuracy: 75.5. Result ID: 698f07893a6c314849f50a5f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"54.1M","FLOPs":"25.0G"}. This record does not assert independent reproduction. ### ConvNeXt-L-384 Top-1 Accuracy: 75.3. Result ID: 698f07893a6c314849f50a6b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"197.8M","FLOPs":"101.0G"}. This record does not assert independent reproduction. ### ViT-L (DeiT III) Top-1 Accuracy: 75.1. Result ID: 698f07893a6c314849f50a70. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"304.4M","FLOPs":"61.6G","Resolution":"224"}. This record does not assert independent reproduction. ### ViT-B↑384 (DeiT III) Top-1 Accuracy: 74.8. Result ID: 698f07893a6c314849f50a6f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"86.9M","FLOPs":"55.5G","Resolution":"384"}. This record does not assert independent reproduction. ### ConvNeXt-B-384 Top-1 Accuracy: 74.7. Result ID: 698f07893a6c314849f50a69. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"88.6M","FLOPs":"45.0G"}. This record does not assert independent reproduction. ### EfficientNetV2-S Top-1 Accuracy: 74. Result ID: 698f07893a6c314849f50a5e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"21.5M","FLOPs":"8.5G"}. This record does not assert independent reproduction. ### ConvNeXt-L-224 Top-1 Accuracy: 74. Result ID: 698f07893a6c314849f50a6a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"197.8M","FLOPs":"34.4G"}. This record does not assert independent reproduction. ### PatchConvNet-B120-224 Top-1 Accuracy: 73.9. Result ID: 698f07893a6c314849f50a67. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"188.6M","FLOPs":"29.9G"}. This record does not assert independent reproduction. ### EfficientNet-B5 Top-1 Accuracy: 73.6. Result ID: 698f07893a6c314849f50a5d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"30.0M","FLOPs":"9.9G"}. This record does not assert independent reproduction. ### ViT-B (DeiT III) Top-1 Accuracy: 73.6. Result ID: 698f07893a6c314849f50a6e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"86.6M","FLOPs":"17.5G","Resolution":"224"}. This record does not assert independent reproduction. ### ConvNeXt-B-224 Top-1 Accuracy: 73.4. Result ID: 698f07893a6c314849f50a68. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"88.6M","FLOPs":"15.4G"}. This record does not assert independent reproduction. ### ViT-S↑384 (DeiT III) Top-1 Accuracy: 73.1. Result ID: 698f07893a6c314849f50a6d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"22.0M","FLOPs":"15.5G","Resolution":"384"}. This record does not assert independent reproduction. ### PatchConvNet-B60-224 Top-1 Accuracy: 72.6. Result ID: 698f07893a6c314849f50a66. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"99.4M","FLOPs":"15.8G"}. This record does not assert independent reproduction. ### RegNetY-16GF Top-1 Accuracy: 72.4. Result ID: 698f07893a6c314849f50a5b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"83.6M","FLOPs":"16.0G"}. This record does not assert independent reproduction. ### EfficientNet-B4 Top-1 Accuracy: 72.3. Result ID: 698f07893a6c314849f50a5c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"19.0M","FLOPs":"4.2G"}. This record does not assert independent reproduction. ### Swin-S-224 Top-1 Accuracy: 71.8. Result ID: 698f07893a6c314849f50a62. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"49.6M","FLOPs":"8.7G"}. This record does not assert independent reproduction. ### RegNetY-8GF Top-1 Accuracy: 71.1. Result ID: 698f07893a6c314849f50a5a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"39.2M","FLOPs":"8.0G"}. This record does not assert independent reproduction. ### PatchConvNet-S60-224 Top-1 Accuracy: 71. Result ID: 698f07893a6c314849f50a65. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"25.2M","FLOPs":"4.0G"}. This record does not assert independent reproduction. ### RegNetY-4GF Top-1 Accuracy: 70.7. Result ID: 698f07893a6c314849f50a59. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"20.6M","FLOPs":"4.0G"}. This record does not assert independent reproduction. ### ResNet-152 Top-1 Accuracy: 70.6. Result ID: 698f07893a6c314849f50a58. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"60.2M","FLOPs":"11.6G"}. This record does not assert independent reproduction. ### ViT-S (DeiT III) Top-1 Accuracy: 70.5. Result ID: 698f07893a6c314849f50a6c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"22.0M","FLOPs":"4.6G","Resolution":"224"}. This record does not assert independent reproduction. ### ResNet-101 Top-1 Accuracy: 70.3. Result ID: 698f07893a6c314849f50a57. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"44.5M","FLOPs":"7.9G"}. This record does not assert independent reproduction. ### Swin-T-224 Top-1 Accuracy: 69.5. Result ID: 698f07893a6c314849f50a61. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"28.3M","FLOPs":"4.5G"}. This record does not assert independent reproduction. ### ResMLP-B24 Top-1 Accuracy: 69. Result ID: 698f07893a6c314849f50a64. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"116.0M","FLOPs":"23.0G"}. This record does not assert independent reproduction. ### ResNet-50 Top-1 Accuracy: 68.7. Result ID: 698f07893a6c314849f50a56. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"nb params":"25.6M","FLOPs":"4.1G"}. This record does not assert independent reproduction. ### Mixer-B/16 Top-1 Accuracy: 63.2. Result ID: 698f07893a6c314849f50a63. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"nb params":"59.9M","FLOPs":"12.6G"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.