List your company

Science Question Answering on ARC Easy (Accuracy & Token Cost)

86.9AccuracyHyPER
79.82881.66483.585.336Feb 6, 2026

Evaluation results

MethodLinks
HyPERModel=DeepSeek-V22026.0286.90.56
HyPERModel=OLMoE2026.0286.50.36
DeepConfModel=DeepSeek-V22026.0286.20.31
DeepConfModel=OLMoE2026.0284.70.25
Self-CertaintyModel=DeepSeek-V22026.0284.51.04
SCModel=DeepSeek-V22026.0284.31
RoEModel=DeepSeek-V22026.0284.31.86
RoEModel=OLMoE2026.0283.82.04
Self-CertaintyModel=OLMoE2026.0282.20.88
SCModel=OLMoE2026.0280.11