List your company

Reasoning on LightEval (test)

0.7452Average Success Rate @8IAC
0.6934080.7068540.72030.733746Jan 22, 2026

Evaluation results

MethodLinks
IACModel=Qwen3-14B2026.010.74520.7728
Reinforce++Model=Qwen3-14B2026.010.73960.7674
IACModel=Qwen3-8B2026.010.73850.7869
DrGRPOModel=Qwen3-14B2026.010.73850.7657
GRPOModel=Qwen3-14B2026.010.73830.7654
RLOOModel=Qwen3-14B2026.010.73640.7636
RLOOModel=Qwen3-8B2026.010.7230.7617
IACModel=Qwen3-4B2026.010.72150.773
Reinforce++Model=Qwen3-8B2026.010.72140.7596
DrGRPOModel=Qwen3-8B2026.010.71560.7536
GRPOModel=Qwen3-8B2026.010.71120.7519
GRPOModel=Qwen3-4B2026.010.70610.7543
RLOOModel=Qwen3-4B2026.010.70240.7539
Reinforce++Model=Qwen3-4B2026.010.70210.7492
DrGRPOModel=Qwen3-4B2026.010.69540.7387