Math Reasoning on GSM8K (Occ, CVaR, RecD)
0.42Occ ScoreFIXED
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| FIXEDInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 0.42 | 0.38 | 210 | |
| RANDOMInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 0.33 | 0.26 | 145 | |
| PERIODICInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 0.25 | 0.22 | 134 | |
| OUTPUT-TRIGGERInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 0.2 | 0.17 | 109 | |
| CONFIDENCE-THRESHOLDInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 0.19 | 0.16 | 112 | |
| RISK PREDICTORInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 0.17 | 0.14 | 98 | |
| VEROICInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 0.13 | 0.11 | 77 |