List your company

Mathematical Reasoning on AIME (pass@1)

66.5Pass@1Uno-Orchestra
-2.6615.29533.2551.205May 6, 2026May 14, 2026May 22, 2026May 30, 2026Jun 7, 2026Jun 15, 2026Jun 23, 2026

Evaluation results

MethodLinks
Uno-Orchestra2026.0566.5
AOrchestra2026.0536.9
xRouter2026.0534.9
AgentOrchestra2026.0531.2
ColdStart-LLM2026.0516.4
SECBackbone=Qwen3-4B-Base, Seeds=12026.0612.9
TACBackbone=Qwen3-4B-Base, Seeds=12026.0612.3
RandomBackbone=Qwen3-4B-Base, Seeds=12026.0610.8
VeriGatePolicy Model=Qwen2.5-7B-Instruct, PRM Model=Qwen-PRM-7B2026.0510
GRPO-VanillaPolicy Model=Qwen2.5-7B-Instruct, PRM Model=Qwen-PRM-7B2026.056.67
SECBackbone=Qwen3-1.7B2026.065.4
RandomBackbone=Qwen3-1.7B2026.065.3
M2OBackbone=Qwen3-1.7B2026.065.2
TACBackbone=Llama3.2-3B2026.065.2
TACBackbone=Qwen3-1.7B2026.065.1
M2OBackbone=Llama3.2-3B2026.064.9
BaseBackbone=Qwen3-1.7B2026.064.3
SECBackbone=Llama3.2-3B2026.063.6
RandomBackbone=Llama3.2-3B2026.063.4
VeriGate (-Gate)Policy Model=Qwen2.5-1.5B-Instruct, PRM Model=Qwen-PRM-7B2026.053.33
PRM-as-ORMPolicy Model=Qwen2.5-7B-Instruct, PRM Model=Qwen-PRM-7B2026.053.33
Dr.GRPOPolicy Model=Qwen2.5-1.5B-Instruct, PRM Model=Qwen-PRM-7B2026.053.3
BaseBackbone=Llama3.2-3B2026.062.2
TACBackbone=Qwen3-0.6B-Base, Seeds=12026.061.7
RandomBackbone=Qwen3-0.6B-Base, Seeds=12026.061.5
SECBackbone=Qwen3-0.6B-Base, Seeds=12026.060.9
GRPO-VanillaPolicy Model=Qwen2.5-1.5B-Instruct, PRM Model=Qwen-PRM-7B2026.050
DAPOPolicy Model=Qwen2.5-1.5B-Instruct, PRM Model=Qwen-PRM-7B2026.050
PRM-as-ORMPolicy Model=Qwen2.5-1.5B-Instruct, PRM Model=Qwen-PRM-7B2026.050
PRM-as-ORM (+Gate)Policy Model=Qwen2.5-1.5B-Instruct, PRM Model=Qwen-PRM-7B2026.050
VeriGate (Simplified)Policy Model=Qwen2.5-1.5B-Instruct, PRM Model=Qwen-PRM-7B2026.050
VeriGatePolicy Model=Qwen2.5-1.5B-Instruct, PRM Model=Qwen-PRM-7B2026.050