List your company

Agent Performance on Tau-bench Telecom

45.39Avg@4 ScoreNoisyAgent
21.67827.83433.9940.146May 26, 2026

Evaluation results

MethodLinks
NoisyAgentbackbone=Qwen3-32B2026.0545.3978.07
GSPObackbone=Qwen3-32B2026.0543.8670.18
DAPObackbone=Qwen3-32B2026.0543.4271.93
GRPObackbone=Qwen3-32B2026.0541.6768.42
NoisyAgentbackbone=Qwen3-8B2026.0540.7964.91
DAPObackbone=Qwen3-8B2026.0539.4763.16
GSPObackbone=Qwen3-8B2026.0539.0461.4
GRPObackbone=Qwen3-8B2026.0537.2857.89
Qwen3-32Bbackbone=Qwen3-32B2026.0528.9549.12
Qwen3-8Bbackbone=Qwen3-8B2026.0522.5942.98