List your company

Math Reasoning on GSM8K, MATH, SVAMP, ASDiv, MAWPS, TAB, MQA, MMLU STEM, and SAT (test)

40.4GSM8K AccuracyBLADE
1.411.52521.6531.775Jun 17, 2026

Evaluation results

MethodLinks
BLADEBackbone=Llama2-7B, Train Toks.=3B, Few-shot CoT=true2026.0640.49.860.866.184.949.627.333.121.943.82,375.1
RHO-1Backbone=Llama2-7B, Train Toks.=3B, Implementation=re-implemented, Few-shot CoT=true2026.06337.657.960.678.948.424.930.128.141.12,140.7
BLADEBackbone=TinyLlama-1.1B, Train Toks.=3B, Few-shot CoT=true2026.0620.88.635.748.966.521.917.222.42529.7494.3
Base-CTBackbone=Llama2-7B, Train Toks.=5B, Few-shot CoT=true2026.0618.14.436.452.766.137.318.533.437.533.81,670.3
RandomBackbone=Llama2-7B, Train Toks.=3B, Few-shot CoT=true2026.0617.35.236.951.964.240.217.132.92532.31,011
MATESBackbone=Llama2-7B, Train Toks.=3B, Sample-level selection=true, Few-shot CoT=true2026.0616.54.435.551.665.738.615.833.638.633.43,274.7
RHO-1Backbone=TinyLlama-1.1B, Train Toks.=3B, Implementation=re-implemented, Few-shot CoT=true2026.0614.64.834.946.761.1171620.815.625.7431.5
BaseBackbone=Llama2-7B, Few-shot CoT=true2026.06143.639.551.763.530.912.432.734.431.4—
MATESBackbone=TinyLlama-1.1B, Train Toks.=3B, Sample-level selection=true, Few-shot CoT=true2026.064.24.817.428.138.31611.913.62517.7831.5
Base-CTBackbone=TinyLlama-1.1B, Train Toks.=5B, Few-shot CoT=true2026.063.73.617.627.83517.19.11818.816.7314.6
RandomBackbone=TinyLlama-1.1B, Train Toks.=3B, Few-shot CoT=true2026.063.63.614.623.730.112.111.418.814.714.7188.8
BaseBackbone=TinyLlama-1.1B, Few-shot CoT=true2026.062.93.21118.120.412.514.616.121.913.4—