List your company

Agent Performance on VitaBench OTA

9.75Avg@4NoisyAgent
1.433.595.757.91May 26, 2026

Evaluation results

MethodLinks
NoisyAgentbackbone=Qwen3-32B2026.059.7515
DAPObackbone=Qwen3-32B2026.059.2515
GSPObackbone=Qwen3-32B2026.05914
GRPObackbone=Qwen3-32B2026.058.7514
Qwen3-32Bbackbone=Qwen3-32B2026.05712
NoisyAgentbackbone=Qwen3-8B2026.0559
GSPObackbone=Qwen3-8B2026.054.58
GRPObackbone=Qwen3-8B2026.054.257
DAPObackbone=Qwen3-8B2026.0547
Qwen3-8Bbackbone=Qwen3-8B2026.051.754