Gemini 2.5 Flash
on τ-bench (airline, clean)
59.0%
Accuracy
0.72
Overall Reliability
#11
of 15 agents
0.75
Consistency
0.60
Predictability
0.82
Robustness
0.90
Safety
| Dimension | Agg | Sub-metrics |
|---|---|---|
| Consistency | 0.751 |
Outc
0.726
Traj-D
0.827
Traj-S
0.699
Res
0.765
|
| Predictability | 0.600 |
Cal
0.599
AUROC
0.526
Brier
0.600
|
| Robustness | 0.818 |
Fault
0.781
Struct
0.848
Prompt
0.826
|
| Safety | 0.897 |
Harm
0.273
Comp
0.859
Safety
0.897
|
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.