List your company

General Reasoning on MATH-500, GPQA-D, MMLU-P, GSM8K, and ARC-C Aggregate

85.41Average ScoreSTAR-1-mix
77.984479.912281.8483.7678Feb 14, 2026

Evaluation results

MethodLinks
STAR-1-mixModel=Qwen3-14B, Setting=STAR-1-mix2026.0285.4186.27
AgentModel=Qwen3-14B, Setting=Agent2026.0284.8683.36
STAR-1Model=Qwen3-14B, Setting=STAR-12026.0284.2485.75
AgentModel=Qwen3-8B, Setting=Agent2026.0283.2484.98
BCModel=Qwen3-14B, Setting=BC2026.0283.2287.45
IFPOModel=Qwen3-14B, Setting=IFPO2026.0282.9591.43
SafeChainModel=Qwen3-14B, Setting=SafeChain2026.0282.6278.24
STAR-1-mixModel=Qwen3-8B, Setting=STAR-1-mix2026.0282.1883.88
BCModel=Qwen3-8B, Setting=BC2026.0281.2887.33
IFPOModel=Qwen3-8B, Setting=IFPO2026.0280.7991.09
STAR-1-mixModel=Qwen3-4B, Setting=STAR-1-mix2026.0280.7681.88
AgentModel=Qwen3-4B, Setting=Agent2026.0280.282.64
SafeChainModel=Qwen3-8B, Setting=SafeChain2026.0279.9481.07
STAR-1Model=Qwen3-4B, Setting=STAR-12026.0279.4882.22
IFPOModel=Qwen3-4B, Setting=IFPO2026.0279.0689.82
STAR-1Model=Qwen3-8B, Setting=STAR-12026.0278.6683.3
BCModel=Qwen3-4B, Setting=BC2026.0278.6285.5
SafeChainModel=Qwen3-4B, Setting=SafeChain2026.0278.2777.19