Math Word Problem Solving on MAWPS (pass@1)
73.7pass@1 AccuracyPOLAR
Evaluation Results
| Method | Links | |
|---|---|---|
| POLARBackbone=LLaMA-3.2-3B-Instruct, Evaluation Protocol=pass@12026.06 | 73.7 | |
| Base (τ=0)Backbone=LLaMA-3.2-3B-Instruct, Temperature=0, Evaluation Protocol=pass@12026.06 | 71.5 | |
| DR.LLMBackbone=LLaMA-3.2-3B-Instruct, Evaluation Protocol=pass@12026.06 | 61.3 | |
| MindSkipBackbone=LLaMA-3.2-3B-Instruct, Evaluation Protocol=pass@12026.06 | 7.5 | |
| ShortGPTBackbone=LLaMA-3.2-3B-Instruct, Evaluation Protocol=pass@12026.06 | 4.6 | |
| FlexiDepthBackbone=LLaMA-3.2-3B-Instruct, Evaluation Protocol=pass@12026.06 | 1.3 |