Machine Unlearning on ToFU GEK (one author (20 samples))
90AccuracyIHL + CE
Evaluation results
| Method | Links | |
|---|---|---|
| IHL + CETraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 90 | |
| IHL + KLTraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 85 | |
| Attention-ShiftingTraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 84 | |
| IHLTraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 64 | |
| GA + CETraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 59 | |
| ULDTraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 47 | |
| NPO + CETraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 45 | |
| NPO + KLTraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 44 | |
| NPOTraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 39 | |
| GA + KLTraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 24 | |
| GATraining Mode=Adapters only, Data Balance=Remaining data equal to TUD2025.10 | 22 |