Behavior Prediction on GROVE Wikipedia 1.0 (P2 & P3)
76Accuracy (Aero.)Adapted CDT
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Adapted CDTAdaptation Study=Dynamic Adaptation, Prediction Model=GPT-4.1, Evaluation Model=GPT-4.12026.04 | 76 | 68.3 | 77.2 | 75 | 78.1 | 72.5 | 64.6 | 68.8 | 68.8 | 72.1 | |
| Retrained CDTAdaptation Study=Complete Retraining, Prediction Model=GPT-4.1, Evaluation Model=GPT-4.12026.04 | 71.2 | 67.5 | 73.4 | 66.7 | 73.4 | 62.5 | 54.2 | 66.7 | 67.2 | 67 | |
| Fixed CDTAdaptation Study=Static/No Adaptation, Prediction Model=GPT-4.1, Evaluation Model=GPT-4.12026.04 | 70.2 | 65 | 69 | 67.7 | 75 | 62.5 | 66.7 | 65.2 | 75.8 | 68.6 |