Human-AI Hybridization on Full Set 60% held-out N=952 (test)
69.3Majority AccuracyBaseline (best)
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Baseline (best)Condition=Baseline, Routing Method=2T/1TC2026.04 | 69.3 | — | 0.4 | — | |
| Top-2 AssistanceCondition=Top-2 Assistance, Routing Method=1TC2026.04 | 69.3 | — | 0.4 | — | |
| Subtask DelegationCondition=Subtask Delegation, Routing Method=1T2026.04 | 69.2 | 69.2 | 0.3 | 0.3 | |
| AI2026.04 | 68.9 | 68.9 | — | — | |
| Human2026.04 | 49.9 | 43.5 | 19 | 25.4 | |
| Baseline (best)Condition=Baseline, Routing Method=2T2026.04 | — | 69.2 | — | 0.3 | |
| Top-2 AssistanceCondition=Top-2 Assistance, Routing Method=1T/1TC2026.04 | — | 69.1 | — | 0.2 |