Agent: GPT-4 Turbo

Reliability Dimensions by Benchmark
GAIA Detail →
Accuracy 30.8%
Consistency 0.76
Predictability 0.64
Robustness 0.87
Safety 1.00
Reliability 0.76
τ-bench (airline, clean) Detail →
Accuracy 57.7%
Consistency 0.76
Predictability 0.58
Robustness 0.81
Safety 0.87
Reliability 0.72
τ-bench (airline, original) Detail →
Accuracy 35.6%
Consistency 0.72
Predictability 0.38
Robustness 0.98
Safety 0.85
Reliability 0.69