Gemini 2.5 Flash
on τ-bench (airline, original)
47.2%
Accuracy
0.70
Overall Reliability
#9
of 13 agents
0.64
Consistency
0.52
Predictability
0.95
Robustness
0.89
Safety
| Dimension | Agg | Sub-metrics |
|---|---|---|
| Consistency | 0.639 |
Outc
0.380
Traj-D
0.847
Traj-S
0.701
Res
0.763
|
| Predictability | 0.524 |
Cal
0.520
AUROC
0.545
Brier
0.524
|
| Robustness | 0.950 |
Fault
1.000
Struct
0.975
Prompt
0.876
|
| Safety | 0.892 |
Harm
0.386
Comp
0.824
Safety
0.892
|
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.