Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Decision classification on Damage-audit 139-case extended (held-out)
Loading...
84.2
Action Accuracy
B1b
73.696
76.423
79.15
81.877
Jun 9, 2026
Action Accuracy
Category Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Action Accuracy
Category Accuracy
B1b
Executor regime=Haiku 4.5
2026.06
84.2
37.4
B1b
Executor regime=Opus 4.6
2026.06
82.7
36.7
Auto-EISR
Executor regime=Haiku...
2026.06
81.3
50.1
Auto-EISR
Executor regime=Opus 4...
2026.06
80.6
50.4
Human-v8
Executor regime=Opus 4.6
2026.06
77.2
48
v1
Executor regime=Haiku 4.5
2026.06
77
35.3
Human-v8
Executor regime=Haiku 4.5
2026.06
75
50
v1
Executor regime=Opus 4.6
2026.06
74.1
36.7
Feedback
Search any
task
Search any
task
