List your company

Mathematical Reasoning on GSM8K (test) (Consensus Metrics)

96.13Accuracy (Raw Consensus)Raw consensus
91.637292.803693.9795.1364May 25, 2026

Evaluation results

MethodLinks
Raw consensusModel=Phi-4, Policy=no-framing majority baseline, Sample count (N)=13192026.0596.13——————
ARBITER-∆Model=Phi-4, Policy=fixed framed+guided parameter-free, Sample count (N)=13192026.0596.1396.440.319624
Raw consensusModel=Qwen3-4B, Policy=no-framing majority baseline, Sample count (N)=13192026.0594.54——————
ARBITER-∆Model=Qwen3-4B, Policy=fixed framed+guided parameter-free, Sample count (N)=13192026.0594.5494.840.316954
Raw consensusModel=Llama-3.1-8B, Policy=no-framing majority baseline, Sample count (N)=13192026.0591.81——————
ARBITER-∆Model=Llama-3.1-8B, Policy=fixed framed+guided parameter-free, Sample count (N)=13192026.0591.8192.650.843520911