Mathematical Reasoning on GSM-Plus (mini)
52.8AccuracyCRESCENT (Llama3-8B-Instruct)
Evaluation Results
| Method | Links | |
|---|---|---|
| CRESCENT (Llama3-8B-Instruct)Training=CRESCENT, Prompting=5-shot2025.02 | 52.8 | |
| Llama3-8B-InstructTraining=Original, Prompting=5-shot2025.02 | 51.2 | |
| CRESCENT (Llama3-8B-Instruct)Training=CRESCENT, Prompting=0-shot2025.02 | 48.6 | |
| Llama3-8B-InstructTraining=Original, Prompting=0-shot2025.02 | 23.1 | |
| CRESCENT (Llama2-7B-Chat)Training=CRESCENT, Prompting=5-shot2025.02 | 14.8 | |
| Llama2-7B-ChatTraining=Original, Prompting=5-shot2025.02 | 13.5 | |
| CRESCENT (Llama2-7B-Chat)Training=CRESCENT, Prompting=0-shot2025.02 | 13 | |
| Llama2-7B-ChatTraining=Original, Prompting=0-shot2025.02 | 11.3 |