# LLM Unlearning on TOFU (Metrics: DF ROUGE, DR ROUGE, ToW, Truth Ratio, MIA) Source: [LLM Unlearning on TOFU (Metrics: DF ROUGE, DR ROUGE, ToW, Truth Ratio, MIA)]() Collection index: [llms.txt]() Data updated: 2026-06-01T08:34:36.347Z Evaluation of LLM Unlearning effectiveness on the TOFU dataset, reporting specific metrics such as DF ROUGE, DR ROUGE, Dtest ROUGE, ToW, Truth Ratio, and MIA. 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 | LLM Unlearning | | Dataset (including any stated split) | TOFU | | Metric | DF ROUGE | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 13 of 13 | | 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 - [DF ROUGE](): Higher is better; unit: PERCENT. - [DR ROUGE](): Higher is better; unit: PERCENT. - [Dtest ROUGE](): Higher is better; unit: PERCENT. - [ToW](): Higher is better; unit: PERCENT. - [Truth Ratio](): Higher is better; unit: PERCENT. - [MIA](): Lower is better; unit: PERCENT. ## Recorded results | Method | DF ROUGE | Result date | Report paper | | --- | --- | --- | --- | | Original | 93.2 | Not recorded | 6a1d11e2dbd7150506f83de5 | | Original | 84.8 | Not recorded | 6a1d11e2dbd7150506f83de5 | | Fine-tune | 72.7 | Not recorded | 6a1d11e2dbd7150506f83de5 | | NPO | 64 | Not recorded | 6a1d11e2dbd7150506f83de5 | | SCRN | 57.3 | Not recorded | 6a1d11e2dbd7150506f83de5 | | SCRUB | 49.8 | Not recorded | 6a1d11e2dbd7150506f83de5 | | DareU | 46.5 | Not recorded | 6a1d11e2dbd7150506f83de5 | | Retraining | 37.5 | Not recorded | 6a1d11e2dbd7150506f83de5 | | Retraining | 34.1 | Not recorded | 6a1d11e2dbd7150506f83de5 | | DareU | 28.8 | Not recorded | 6a1d11e2dbd7150506f83de5 | | GDiff | 28.1 | Not recorded | 6a1d11e2dbd7150506f83de5 | | IDK | 19.1 | Not recorded | 6a1d11e2dbd7150506f83de5 | | GA | 0.3 | Not recorded | 6a1d11e2dbd7150506f83de5 | ### Original DF ROUGE: 93.2. Result ID: 6a1d43c1fbc373592c1677b3. [Reporting paper]() Recorded evaluation configuration: {"Full-parameter unlearning":"true","LoRA":"false","Backbone":"Llama2-7B","Seed":"42"}. This record does not assert independent reproduction. ### Original DF ROUGE: 84.8. Result ID: 6a1d43c1fbc373592c1677a0. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### Fine-tune DF ROUGE: 72.7. Result ID: 6a1d43c1fbc373592c1677a2. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### NPO DF ROUGE: 64. Result ID: 6a1d43c1fbc373592c1677a8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### SCRN DF ROUGE: 57.3. Result ID: 6a1d43c1fbc373592c1677a6. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### SCRUB DF ROUGE: 49.8. Result ID: 6a1d43c1fbc373592c1677a5. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### DareU DF ROUGE: 46.5. Result ID: 6a1d43c1fbc373592c1677a9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### Retraining DF ROUGE: 37.5. Result ID: 6a1d43c1fbc373592c1677a1. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### Retraining DF ROUGE: 34.1. Result ID: 6a1d43c1fbc373592c1677b4. [Reporting paper]() Recorded evaluation configuration: {"Full-parameter unlearning":"true","LoRA":"false","Backbone":"Llama2-7B","Seed":"42"}. This record does not assert independent reproduction. ### DareU DF ROUGE: 28.8. Result ID: 6a1d43c1fbc373592c1677b5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Full-parameter unlearning":"true","LoRA":"false","Backbone":"Llama2-7B","Seed":"42"}. This record does not assert independent reproduction. ### GDiff DF ROUGE: 28.1. Result ID: 6a1d43c1fbc373592c1677a4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### IDK DF ROUGE: 19.1. Result ID: 6a1d43c1fbc373592c1677a7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. ### GA DF ROUGE: 0.3. Result ID: 6a1d43c1fbc373592c1677a3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.