Mathematical Reasoning on OlympiadBench Math (Relative Cost, Token Consumption)
1Relative CostCoT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CoTEvaluation Model=GPT-4o-mini2026.03 | 1 | 4,662 | |
| Analogical PromptingEvaluation Model=GPT-4o-mini2026.03 | 1.1 | — | |
| Role / Persona PromptingEvaluation Model=GPT-4o-mini2026.03 | 1.21 | — | |
| Prompt CanvasEvaluation Model=GPT-4o-mini2026.03 | 1.21 | — | |
| TDA-RCEvaluation Model=GPT-4o-mini2026.03 | 1.21 | — | |
| Instruction InductionEvaluation Model=GPT-4o-mini2026.03 | 1.23 | — | |
| HoTEvaluation Model=GPT-4o-mini2026.03 | 1.36 | — | |
| Self-RefineEvaluation Model=GPT-4o-mini2026.03 | 2.82 | — | |
| AoTEvaluation Model=GPT-4o-mini2026.03 | 3.86 | — | |
| CoT-SCEvaluation Model=GPT-4o-mini, n=52026.03 | 5.07 | — | |
| AFlowEvaluation Model=GPT-4o-mini2026.03 | 5.08 | — | |
| ToTEvaluation Model=GPT-4o-mini2026.03 | 7.03 | — | |
| GoTEvaluation Model=GPT-4o-mini2026.03 | 7.46 | — | |
| FoTEvaluation Model=GPT-4o-mini, n=82026.03 | 8.09 | — |