Mathematical Reasoning on GSM8K (Relative Cost, Token Consumption)
1Relative CostCoT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CoTEvaluation Model=GPT-4o-mini2026.03 | 1 | 264 | |
| Analogical PromptingEvaluation Model=GPT-4o-mini2026.03 | 1.05 | — | |
| TDA-RCEvaluation Model=GPT-4o-mini2026.03 | 1.09 | — | |
| Role / Persona PromptingEvaluation Model=GPT-4o-mini2026.03 | 1.1 | — | |
| Prompt CanvasEvaluation Model=GPT-4o-mini2026.03 | 1.11 | — | |
| Instruction InductionEvaluation Model=GPT-4o-mini2026.03 | 1.12 | — | |
| HoTEvaluation Model=GPT-4o-mini2026.03 | 1.17 | — | |
| Self-RefineEvaluation Model=GPT-4o-mini2026.03 | 2.37 | — | |
| AoTEvaluation Model=GPT-4o-mini2026.03 | 3.29 | — | |
| AFlowEvaluation Model=GPT-4o-mini2026.03 | 3.74 | — | |
| ToTEvaluation Model=GPT-4o-mini2026.03 | 4.69 | — | |
| CoT-SCEvaluation Model=GPT-4o-mini, n=52026.03 | 4.93 | — | |
| GoTEvaluation Model=GPT-4o-mini2026.03 | 4.97 | — | |
| FoTEvaluation Model=GPT-4o-mini, n=82026.03 | 8.02 | — |