Continual Learning on Sequential 5 CL Tasks RS, Med, AD, Sci, Fin
78.8RS AccuracyModel-Dowser
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Model-DowserBackbone=LLaVA-v1.5-7B2026.02 | 78.8 | 61.2 | 48 | 53.5 | 91.4 | 66.6 | 69.1 | 66.6 | 69.6 | -2.5 | 63.8 | |
| MoeLORABackbone=LLaVA-v1.5-7B2026.02 | 73.2 | 51.5 | 35 | 49.6 | 90.9 | 60 | 67.9 | 60 | 64.8 | -7.8 | 60.6 | |
| TailorBackbone=LLaVA-v1.5-7B2026.02 | 66 | 50 | 21.2 | 51.1 | 91.3 | 55.9 | 69.1 | 55.9 | 66.6 | -13.2 | 47 | |
| Full-FTBackbone=LLaVA-v1.5-7B2026.02 | 60 | 50.1 | 20.6 | 51 | 91.6 | 54.7 | 70.2 | 54.7 | 66.2 | -15.6 | 46.3 | |
| Zero-shotBackbone=LLaVA-v1.5-7B2026.02 | 32.3 | 35.5 | 15.6 | 42.3 | 62.5 | 37.6 | — | — | — | — | — |