Future navigation-style prediction on PowerWash (cross-validation)
61.6Accuracy (20 min)PLVM
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| PLVMModel family=Process-Level Latent Variable Model2026.05 | 61.6 | 65.3 | 65.2 | 0.64 | |
| Best Single-task (Bungalow)Model family=Single-task process model, Observation source=Bungalow2026.05 | 57.9 | 62.9 | 63.6 | 0.62 | |
| Best OutcomeModel family=Outcome baseline2026.05 | 56 | 53 | 52.5 | 0.54 |