Question Answering on HotpotQA (Occ, CVaR, RecD)
52OccFIXED
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| FIXEDInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 52 | 48 | 276 | |
| RANDOMInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 45 | 37 | 216 | |
| PERIODICInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 39 | 30 | 193 | |
| OUTPUT-TRIGGERInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 31 | 27 | 168 | |
| CONFIDENCE-THRESHOLDInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 31 | 26 | 174 | |
| RISK PREDICTORInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 27 | 23 | 158 | |
| VEROICInference Path=LLaMA-3.1-8B -> LLaMA-3.1-70B, Setting=CMIS2026.04 | 22 | 19 | 128 |