AnnoMI
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
AnnoMI (80/20)
52.6Macro F1 Score
24
AnnoMI session-level
1.28R/Q
17
AnnoMI high-quality subset
1.28R/Q
12
AnnoMI (test)
2.85Cultivate Score
12
AnnoMI (full)
3.4MI-i
11
AnnoMI
2.81Exploration Focus
10
AnnoMI (test)
4.45Client Readability
5
AnnoMI (test)
9.01PE
5
AnnoMI (test)
2.1MI-i (%)
5
AnnoMI Expert Evaluation Subset
4.06Cultivating Change Talk
4
AnnoMI 1.0 (test)
4.72Personas
4
AnnoMI v1 (test)
5.5MI-i Score (%)
4
AnnoMI
5Avg Receptivity (Threshold 1.0)
2
AnnoMI (human evaluation)
21RAND Score
2