Continual Learning on 5-Dataset
93.56AccuracyTRGP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TRGPNetwork=multi-head, Runs=5-run average2022.11 | 93.56 | -0.04 | |
| CUBERNetwork=multi-head, Runs=5-run average, epsilon_2=0.02022.11 | 93.48 | 0 | |
| MultitaskNetwork=multi-head, Runs=5-run average2022.11 | 91.54 | — | |
| HATNetwork=multi-head, Runs=5-run average2022.11 | 91.32 | -1 | |
| GPMNetwork=multi-head, Runs=5-run average2022.11 | 91.22 | -1 | |
| EWCNetwork=multi-head, Runs=5-run average2022.11 | 88.64 | -4 | |
| ER ResNetwork=multi-head, Runs=5-run average2022.11 | 88.31 | -4 | |
| A-GEMNetwork=multi-head, Runs=5-run average2022.11 | 84.04 | -12 | |
| InfluenceCL2023.04 | 61.6 | -0.05 | |
| OCS2023.04 | 61.5 | -0.03 | |
| GSS2023.04 | 60.2 | -0.07 | |
| Vanilla IF2023.04 | 59.7 | -0.07 | |
| Bilevel2023.04 | 58.1 | -0.08 | |
| ER-MIR2023.04 | 56.9 | -0.11 | |
| Stable SGD2023.04 | 53.4 | -0.16 | |
| EWC2023.04 | 42.7 | -0.28 |