# Machine Unlearning on TOFU 10% Forget Set (v1.0) Source: [Machine Unlearning on TOFU 10% Forget Set (v1.0)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:44:59.191Z Evaluation of Machine Unlearning effectiveness on the 10% forget set subset of the TOFU 1.0 dataset, reporting metrics such as FQ, MU, and ES scores. 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) | TOFU 10% forget set 1.0 | | Metric | FQ | | Unit | SCALAR | | Direction | Lower is better | | Results loaded | 18 of 18 | | 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 - [FQ](): Lower is better; unit: SCALAR. - [MU](): Higher is better; unit: PERCENT. - [ES(Df)](): Lower is better; unit: PERCENT. - [ES(Dr)](): Lower is better; unit: PERCENT. - [Gib](): Higher is better; unit: SCALAR. ## Recorded results | Method | FQ | Result date | Report paper | | --- | --- | --- | --- | | GA | -238.973 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | GA | -236.07 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | GD | -26.8 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | Full | -26.444 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | Full | -21.408 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | GD | -15.484 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | TV | -12.506 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | NPO | -10.244 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | TV | -10.22 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | NPO+ | -7.042 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | NPO | -4.59 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | NPO+ | -4.481 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | PerTA-fisher | -2.99 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | PerTA-grad | -1.708 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | PerTA-fisher | -1.686 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | PerTA-grad | -1.107 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | GT | 0 | Not recorded | 697c3b76a8b01e76cc8da7b2 | | GT | 0 | Not recorded | 697c3b76a8b01e76cc8da7b2 | ### GA FQ: -238.973. Result ID: 6998f6125698665ec45b1bd4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1B","Category":"Training-based"}. This record does not assert independent reproduction. ### GA FQ: -236.07. Result ID: 6998f6125698665ec45b1bdd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B","Category":"Training-based"}. This record does not assert independent reproduction. ### GD FQ: -26.8. Result ID: 6998f6125698665ec45b1bde. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B","Category":"Training-based"}. This record does not assert independent reproduction. ### Full FQ: -26.444. Result ID: 6998f6125698665ec45b1bdb. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"3B","Status":"Before unlearning"}. This record does not assert independent reproduction. ### Full FQ: -21.408. Result ID: 6998f6125698665ec45b1bd2. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"1B","Status":"Before unlearning"}. This record does not assert independent reproduction. ### GD FQ: -15.484. Result ID: 6998f6125698665ec45b1bd5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1B","Category":"Training-based"}. This record does not assert independent reproduction. ### TV FQ: -12.506. Result ID: 6998f6125698665ec45b1bd8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1B","Category":"Task Arithmetic-based"}. This record does not assert independent reproduction. ### NPO FQ: -10.244. Result ID: 6998f6125698665ec45b1bd6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1B","Category":"Training-based"}. This record does not assert independent reproduction. ### TV FQ: -10.22. Result ID: 6998f6125698665ec45b1be1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B","Category":"Task Arithmetic-based"}. This record does not assert independent reproduction. ### NPO+ FQ: -7.042. Result ID: 6998f6125698665ec45b1be0. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"3B","Category":"Training-based"}. This record does not assert independent reproduction. ### NPO FQ: -4.59. Result ID: 6998f6125698665ec45b1bdf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B","Category":"Training-based"}. This record does not assert independent reproduction. ### NPO+ FQ: -4.481. Result ID: 6998f6125698665ec45b1bd7. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"1B","Category":"Training-based"}. This record does not assert independent reproduction. ### PerTA-fisher FQ: -2.99. Result ID: 6998f6125698665ec45b1be3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B","Category":"Task Arithmetic-based"}. This record does not assert independent reproduction. ### PerTA-grad FQ: -1.708. Result ID: 6998f6125698665ec45b1be2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"3B","Category":"Task Arithmetic-based"}. This record does not assert independent reproduction. ### PerTA-fisher FQ: -1.686. Result ID: 6998f6125698665ec45b1bda. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1B","Category":"Task Arithmetic-based"}. This record does not assert independent reproduction. ### PerTA-grad FQ: -1.107. Result ID: 6998f6125698665ec45b1bd9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"1B","Category":"Task Arithmetic-based"}. This record does not assert independent reproduction. ### GT FQ: 0. Result ID: 6998f6125698665ec45b1bd3. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"1B","Status":"Ground truth model"}. This record does not assert independent reproduction. ### GT FQ: 0. Result ID: 6998f6125698665ec45b1bdc. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"3B","Status":"Ground truth model"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.