Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Human-AI Hybridization on Full Set 60% held-out N=952 (test)
Loading...
69.3
Majority Accuracy
Baseline (best)
49.124
54.362
59.6
64.838
Apr 13, 2026
Majority Accuracy
Individual Accuracy
Majority Accuracy Improvement vs AI (pp)
Individual Accuracy Improvement vs AI (pp)
Updated 2mo ago
Evaluation Results
Method
Method
Links
Majority Accuracy
Individual Accuracy
Majority Accuracy Improvement vs AI (pp)
Individual Accuracy Improvement vs AI (pp)
Baseline (best)
Condition=Baseline, Ro...
2026.04
69.3
-
0.4
-
Top-2 Assistance
Condition=Top-2 Assist...
2026.04
69.3
-
0.4
-
Subtask Delegation
Condition=Subtask Dele...
2026.04
69.2
69.2
0.3
0.3
AI
2026.04
68.9
68.9
-
-
Human
2026.04
49.9
43.5
19
25.4
Baseline (best)
Condition=Baseline, Ro...
2026.04
-
69.2
-
0.3
Top-2 Assistance
Condition=Top-2 Assist...
2026.04
-
69.1
-
0.2
Feedback
Search any
task
Search any
task