# Image Classification on Tiny-ImageNet Bucket 20 Retain Source: [Image Classification on Tiny-ImageNet Bucket 20 Retain]() Collection index: [llms.txt]() Data updated: 2026-04-10T09:24:36.449Z Evaluation of Image Classification performance on the specific 'Bucket 20 Retain' subset of the Tiny-ImageNet dataset, reporting 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) | Tiny-ImageNet Bucket 20 Retain | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 15 of 15 | | 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 - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Original | 84.87 | Not recorded | 69d884367d0172140ec77ba4 | | Original | 84.81 | Not recorded | 69d884367d0172140ec77ba4 | | Retain-only Retain | 84.45 | Not recorded | 69d884367d0172140ec77ba4 | | Retain-only Retain | 84.14 | Not recorded | 69d884367d0172140ec77ba4 | | Original | 83.49 | Not recorded | 69d884367d0172140ec77ba4 | | Retain-only Retain | 82.74 | Not recorded | 69d884367d0172140ec77ba4 | | Salun with CMF | 81.36 | Not recorded | 69d884367d0172140ec77ba4 | | Random-label with CMF | 80.57 | Not recorded | 69d884367d0172140ec77ba4 | | Salun with CMF | 78.34 | Not recorded | 69d884367d0172140ec77ba4 | | Random-label with CMF | 78.12 | Not recorded | 69d884367d0172140ec77ba4 | | Salun with CMF | 77.84 | Not recorded | 69d884367d0172140ec77ba4 | | Random-label with CMF | 77.55 | Not recorded | 69d884367d0172140ec77ba4 | | NegGrad+ with CMF | 69.65 | Not recorded | 69d884367d0172140ec77ba4 | | NegGrad+ with CMF | 65.38 | Not recorded | 69d884367d0172140ec77ba4 | | NegGrad+ with CMF | 63.65 | Not recorded | 69d884367d0172140ec77ba4 | ### Original Accuracy: 84.87. Result ID: 69d8c1cc47f34ce924c1f18b. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Head":"Linear Probe","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Original Accuracy: 84.81. Result ID: 69d8c1cc47f34ce924c1f18a. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Head":"Output","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Retain-only Retain Accuracy: 84.45. Result ID: 69d8c1cc47f34ce924c1f18e. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Head":"Linear Probe","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Retain-only Retain Accuracy: 84.14. Result ID: 69d8c1cc47f34ce924c1f18d. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Head":"Output","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Original Accuracy: 83.49. Result ID: 69d8c1cc47f34ce924c1f18c. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Head":"NCC","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Retain-only Retain Accuracy: 82.74. Result ID: 69d8c1cc47f34ce924c1f18f. [Reporting paper]() Recorded evaluation configuration: {"Evaluation Head":"NCC","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Salun with CMF Accuracy: 81.36. Result ID: 69d8c1cc47f34ce924c1f194. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"Linear Probe","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Random-label with CMF Accuracy: 80.57. Result ID: 69d8c1cc47f34ce924c1f191. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"Linear Probe","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Salun with CMF Accuracy: 78.34. Result ID: 69d8c1cc47f34ce924c1f193. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"Output","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Random-label with CMF Accuracy: 78.12. Result ID: 69d8c1cc47f34ce924c1f190. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"Output","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Salun with CMF Accuracy: 77.84. Result ID: 69d8c1cc47f34ce924c1f195. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"NCC","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### Random-label with CMF Accuracy: 77.55. Result ID: 69d8c1cc47f34ce924c1f192. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"NCC","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### NegGrad+ with CMF Accuracy: 69.65. Result ID: 69d8c1cc47f34ce924c1f197. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"Linear Probe","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### NegGrad+ with CMF Accuracy: 65.38. Result ID: 69d8c1cc47f34ce924c1f196. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"Output","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. ### NegGrad+ with CMF Accuracy: 63.65. Result ID: 69d8c1cc47f34ce924c1f198. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Evaluation Head":"NCC","Backbone":"ViT-S/16","Pre-training":"ImageNet"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.