List your company

Mathematical Reasoning on GSM8K FLORES-200 (10 low-resourced languages) (test)

67.93Kazakh (Cyrl) AccuracyDIP
12.820427.127741.43555.7423Nov 2, 2024

Evaluation results

MethodLinks
DIPModel=GPT-4o, English Pivoting (EP)=true, Chain-of-thought (CT)=true2024.1167.9346.1780.3667.153.343.2968.6160.563.6868.3161.92
English Pivot ThoughtModel=GPT-4o, English Pivoting (EP)=true, Chain-of-thought (CT)=true2024.1161.1136.7760.3557.1649.1322.6760.9620.8544.2835.4844.88
English PivotingModel=GPT-4o, English Pivoting (EP)=true, Chain-of-thought (CT)=false2024.1123.6514.0319.9419.7919.2610.3121.919.5517.8217.4417.37
DIP w/ EP w/o CTModel=GPT-4o, English Pivoting (EP)=true, Chain-of-thought (CT)=false2024.1123.4316.324.9421.2318.514.8623.5819.1822.1422.3720.65
DIP w/o EP w/o CTModel=GPT-4o, English Pivoting (EP)=false, Chain-of-thought (CT)=false2024.1120.3912.0522.0615.4713.5712.4319.0313.9516.3818.216.35
Standard PromptingModel=GPT-4o, English Pivoting (EP)=false, Chain-of-thought (CT)=false2024.1117.899.115.6211.5211.35.9114.255.849.559.411.04
Non-insertion PromptingModel=GPT-4o2024.1114.946.2913.8710.0810.246.3712.897.0510.248.1910.02