Automated Short Answer Scoring on SAS History subject (OOD)
47.7AccuracyEDIT-RL
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| EDIT-RLCategory=EDIT (Ours)2026.06 | 47.7 | 62.3 | 1.615 | 89.9 | |
| GRPOCategory=RL-only2026.06 | 47.1 | 59.7 | 1.883 | 87.3 | |
| DGPOCategory=RL-only2026.06 | 46.2 | 60 | 1.865 | 87.6 | |
| RLCategory=InT2026.06 | 46.1 | 62.5 | 1.569 | 88.9 | |
| EDIT-SFTCategory=EDIT (Ours)2026.06 | 45.4 | 56.9 | 1.952 | 86.3 | |
| SFTCategory=InT2026.06 | 42.6 | 56.6 | 1.852 | 87.1 | |
| BaseCategory=Base2026.06 | 41.9 | 51.2 | 2.612 | 78.3 |