# Machine Unlearning on TinyImageNet (50% Random Removal) Source: [Machine Unlearning on TinyImageNet (50% Random Removal)]() Collection index: [llms.txt]() Data updated: 2026-05-19T10:22:55.952Z Evaluation of Machine Unlearning performance on the TinyImageNet dataset configured for 50% random data removal. Metrics include accuracy retention, unlearning efficacy, MIA results, and computational cost. 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 | Machine Unlearning | | Dataset (including any stated split) | TinyImageNet 50% random data removal | | Metric | Retain Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 7 of 7 | | 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 - [Retain Accuracy](): Higher is better; unit: PERCENT. - [Unlearn Accuracy](): Higher is better; unit: PERCENT. - [Test Accuracy](): Higher is better; unit: PERCENT. - [MIA Efficacy](): Higher is better; unit: PERCENT. - [Average Gap](): Higher is better; unit: SCALAR. - [Computational Cost (PFLOPs)](): Lower is better; unit: SIZE. ## Recorded results | Method | Retain Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Retrain | 99.99 | Not recorded | 68fa7c441e209131cea8d47a | | FT | 99.99 | Not recorded | 68fa7c441e209131cea8d47a | | NegGrad+ | 99.99 | Not recorded | 68fa7c441e209131cea8d47a | | ℓ1-sparse | 99.99 | Not recorded | 68fa7c441e209131cea8d47a | | NoT | 99.99 | Not recorded | 68fa7c441e209131cea8d47a | | CoUn | 99.98 | Not recorded | 68fa7c441e209131cea8d47a | | SalUn | 93.59 | Not recorded | 68fa7c441e209131cea8d47a | ### Retrain Retain Accuracy: 99.99. Result ID: 6a0c39ebfbc373592c02f05f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-18"}. This record does not assert independent reproduction. ### FT Retain Accuracy: 99.99. Result ID: 6a0c39ebfbc373592c02f060. [Reporting paper]() Recorded evaluation configuration: {"Model":"ResNet-18"}. This record does not assert independent reproduction. ### NegGrad+ Retain Accuracy: 99.99. Result ID: 6a0c39ebfbc373592c02f061. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"ResNet-18"}. This record does not assert independent reproduction. ### ℓ1-sparse Retain Accuracy: 99.99. Result ID: 6a0c39ebfbc373592c02f062. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-18"}. This record does not assert independent reproduction. ### NoT Retain Accuracy: 99.99. Result ID: 6a0c39ebfbc373592c02f064. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-18"}. This record does not assert independent reproduction. ### CoUn Retain Accuracy: 99.98. Result ID: 6a0c39ebfbc373592c02f065. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-18"}. This record does not assert independent reproduction. ### SalUn Retain Accuracy: 93.59. Result ID: 6a0c39ebfbc373592c02f063. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"ResNet-18"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.