Claude Sonnet 4
on τ-bench (airline, clean)
78.2%
Accuracy
0.86
Overall Reliability
#4
of 15 agents
0.76
Consistency
0.85
Predictability
0.96
Robustness
0.96
Safety
| Dimension | Agg | Sub-metrics |
|---|---|---|
| Consistency | 0.763 |
Outc
0.624
Traj-D
0.861
Traj-S
0.785
Res
0.843
|
| Predictability | 0.853 |
Cal
0.896
AUROC
0.666
Brier
0.853
|
| Robustness | 0.963 |
Fault
0.989
Struct
0.984
Prompt
0.918
|
| Safety | 0.955 |
Harm
0.417
Comp
0.923
Safety
0.955
|
Per-Task Outcome Consistency
Each cell represents a task. Color shows outcome consistency across runs. Hover to see task ID.
Consistency Distribution
KDE of per-task outcome consistency. Peaks at 0 or 1 indicate polarized behavior.
Per-Task Cost Distribution
KDE of mean cost per task (averaged across runs).
Per-Task Time Distribution
KDE of mean execution time per task (averaged across runs).
Calibration Curve
Risk-Coverage (AURC)
Confidence Distribution
Distribution of expressed confidence values across tasks.