Agent objective detection on trajectories Phase I (held-out)
100AccuracyQBM (UCIP)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| QBM (UCIP)Protocol=UCIP, Sample size (n)=30 per class2026.03 | 100 | 100 | 0.381 | |
| AutoencoderSample size (n)=30 per class2026.03 | 71.7 | 64.6 | 0.001 | |
| Classical RBMΓ=0, Sample size (n)=30 per class2026.03 | 50 | 0 | -0.083 |