List your company

Math Reasoning on SVAMP (Multilingual Performance Breakdown)

78.33Kazakh (Cyrl) AccuracyDIP
51.643658.571865.572.4282Nov 2, 2024

Evaluation results

MethodLinks
DIPBackbone=GPT-4o, EP=true, CT=true2024.1178.335789.6776.6765.3365.6777.677178.677573.5
DIP w/ EP w/o CTBackbone=GPT-4o, EP=true, CT=false2024.1166.6755.6775.6773.675755.67706267.336664.97
DIP w/o EP w/o CTBackbone=GPT-4o, EP=false, CT=false2024.1166437365.674945.3362.6748.6762.3361.6757.73
English Pivot ThoughtBackbone=GPT-4o2024.1161.3361.3358.677160.6741.677136.6769.3355.3358.7
English PivotingBackbone=GPT-4o2024.1154.6751.3351.676353.6740.3363.673557.3351.6752.23
Standard PromptingBackbone=GPT-4o2024.1153.334246.3348.3336.6728.335325.3337.6740.6741.17
Non-insertion PromptingBackbone=GPT-4o2024.1152.6734.33454638.3329.675328.6745.6741.3341.47