List your company

Agent Performance on VitaBench Delivery

29Avg@4NoisyAgent
13.1417.257521.37525.4925May 26, 2026

Evaluation results

MethodLinks
NoisyAgentbackbone=Qwen3-32B2026.052943
DAPObackbone=Qwen3-32B2026.0527.7541
GSPObackbone=Qwen3-32B2026.0527.2540
GRPObackbone=Qwen3-32B2026.052740
Qwen3-32Bbackbone=Qwen3-32B2026.052335
NoisyAgentbackbone=Qwen3-8B2026.0522.2535
DAPObackbone=Qwen3-8B2026.0521.534
GSPObackbone=Qwen3-8B2026.0521.2533
GRPObackbone=Qwen3-8B2026.052133
Qwen3-8Bbackbone=Qwen3-8B2026.0513.7522