# Image Classification on MNIST (test) (Natural and Certified Accuracy) Source: [Image Classification on MNIST (test) (Natural and Certified Accuracy)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:53:33.240Z Evaluation of Image Classification performance on the MNIST test set, reporting both Natural Accuracy and Certified 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) | MNIST full (test) | | Metric | Natural Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 12 of 12 | | 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 - [Natural Accuracy](): Higher is better; unit: PERCENT. - [Certified Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Natural Accuracy | Result date | Report paper | | --- | --- | --- | --- | | SABR | 99.23 | Not recorded | 68fa71081e209131cea4ec2a | | COLT | 99.2 | Not recorded | 68fa71081e209131cea4ec2a | | TAPS | 99.19 | Not recorded | 68fa71081e209131cea4ec2a | | STAPS | 99.15 | Not recorded | 68fa71081e209131cea4ec2a | | SORTNET | 99.01 | Not recorded | 68fa71081e209131cea4ec2a | | IBP | 98.84 | Not recorded | 68fa71081e209131cea4ec2a | | SABR | 98.75 | Not recorded | 68fa71081e209131cea4ec2a | | STAPS | 98.53 | Not recorded | 68fa71081e209131cea4ec2a | | SORTNET | 98.46 | Not recorded | 68fa71081e209131cea4ec2a | | TAPS | 97.94 | Not recorded | 68fa71081e209131cea4ec2a | | IBP | 97.67 | Not recorded | 68fa71081e209131cea4ec2a | | COLT | 97.3 | Not recorded | 68fa71081e209131cea4ec2a | ### SABR Natural Accuracy: 99.23. Result ID: 699172df3a6c314849083f04. [Reporting paper]() Recorded evaluation configuration: {"epsilon\_infinity":"0.1"}. This record does not assert independent reproduction. ### COLT Natural Accuracy: 99.2. Result ID: 699172df3a6c314849083f01. [Reporting paper]() Recorded evaluation configuration: {"epsilon\_infinity":"0.1"}. This record does not assert independent reproduction. ### TAPS Natural Accuracy: 99.19. Result ID: 699172df3a6c314849083f05. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"epsilon\_infinity":"0.1"}. This record does not assert independent reproduction. ### STAPS Natural Accuracy: 99.15. Result ID: 699172df3a6c314849083f06. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"epsilon\_infinity":"0.1"}. This record does not assert independent reproduction. ### SORTNET Natural Accuracy: 99.01. Result ID: 699172df3a6c314849083f03. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon\_infinity":"0.1"}. This record does not assert independent reproduction. ### IBP Natural Accuracy: 98.84. Result ID: 699172df3a6c314849083f02. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"epsilon\_infinity":"0.1"}. This record does not assert independent reproduction. ### SABR Natural Accuracy: 98.75. Result ID: 699172df3a6c314849083f0a. [Reporting paper]() Recorded evaluation configuration: {"epsilon\_infinity":"0.3"}. This record does not assert independent reproduction. ### STAPS Natural Accuracy: 98.53. Result ID: 699172df3a6c314849083f0c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"epsilon\_infinity":"0.3"}. This record does not assert independent reproduction. ### SORTNET Natural Accuracy: 98.46. Result ID: 699172df3a6c314849083f09. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon\_infinity":"0.3"}. This record does not assert independent reproduction. ### TAPS Natural Accuracy: 97.94. Result ID: 699172df3a6c314849083f0b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"epsilon\_infinity":"0.3"}. This record does not assert independent reproduction. ### IBP Natural Accuracy: 97.67. Result ID: 699172df3a6c314849083f08. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"epsilon\_infinity":"0.3"}. This record does not assert independent reproduction. ### COLT Natural Accuracy: 97.3. Result ID: 699172df3a6c314849083f07. [Reporting paper]() Recorded evaluation configuration: {"epsilon\_infinity":"0.3"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.