Automated Essay Scoring on ASAP Prompt 3 All essays (val)
0.756Quadratic Weighted Kappa (QWK)NPCR
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| NPCR2026.04 | 0.756 | — | — | — | — | |
| BERTarchitecture=Standard BERT, validation=5-fold CV2026.04 | 0.647 | — | — | — | 0.648 | |
| Human-human QWK2026.04 | — | — | — | 0.769 | — | |
| Human-like ceiling2026.04 | — | — | 0.818 | — | — | |
| Theoretical ceiling2026.04 | — | 0.932 | — | — | — |