Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Agentic Reasoning and Task Execution on Controlled-domain Coding
Loading...
0.812
Mean Task Rubric Score (G)
Offline AW Controller
0.708
0.735
0.762
0.789
Jul 5, 2026
Mean Task Rubric Score (G)
Change in Mean Task Rubric Score (Delta G)
Change in HMS Metric (Delta HMS)
95% Confidence Interval (Lower Bound)
P-value
CheckBeforeSubmit Rate (CBS)
Updated 18d ago
Evaluation Results
Method
Method
Links
Mean Task Rubric Score (G)
Change in Mean Task Rubric Score (Delta G)
Change in HMS Metric (Delta HMS)
95% Confidence Interval (Lower Bound)
P-value
CheckBeforeSubmit Rate (CBS)
Offline AW Controller
Policy Variant=AW, See...
2026.07
0.812
0.1
0.054
0.059
0.001
0.178
Base
Policy Variant=Base, S...
2026.07
0.712
-
-
-
-
-
Feedback
Search any
task
Search any
task