Mathematical Reasoning on AMC12 (test)
46AccuracyRIOT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| RIOTOptimization Type=Automatic Prompt Optimization2025.06 | 46 | 5.1 | |
| APEOptimization Type=Automatic Prompt Optimization2025.06 | 42.8 | 1.9 | |
| OPROOptimization Type=Automatic Prompt Optimization2025.06 | 42.8 | 1.9 | |
| DSPyOptimization Type=Automatic Prompt Optimization2025.06 | 41.9 | 1 | |
| Zero-Shot CoTPrompting Strategy=Chain of Thought, Few-shot shots=02025.06 | 40.9 | 0 | |
| TextGradOptimization Type=Automatic Prompt Optimization2025.06 | 40.9 | 0 | |
| Four-Shot CoTPrompting Strategy=Chain of Thought, Few-shot shots=42025.06 | 37.2 | -3.7 | |
| Twenty-Shot CoTPrompting Strategy=Chain of Thought, Few-shot shots=202025.06 | 30.7 | -10.2 |