Mathematical Reasoning on MathBench Arithmetic
75AccuracyVanilla CoT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Vanilla CoTPrompting Strategy=Vanilla CoT, Model=GPT-4o-mini2024.12 | 75 | 313.51 | 78.58 | |
| TALE-EPPrompting Strategy=TALE-EP, Model=GPT-4o-mini2024.12 | 73.67 | 39.6 | 18.62 | |
| Directly AnsweringPrompting Strategy=Directly Answering, Model=GPT-4o-mini2024.12 | 59.67 | 41.1 | 9.78 |