Selective Generation on CoQA (PRR (AlignScore))
47.2PRR (AlignScore)SentenceSAR
Evaluation Results
| Method | Links | |
|---|---|---|
| SentenceSARModel=Llama 8b v3.12025.02 | 47.2 | |
| SARModel=Llama 8b v3.12025.02 | 46.5 | |
| Maximum Sequence ProbabilityModel=Llama 8b v3.12025.02 | 45 | |
| PerplexityModel=Llama 8b v3.12025.02 | 45 | |
| HUQ-SATRMDModel=Llama 8b v3.12025.02 | 45 | |
| EigValLaplacian NLI Score Entail.Model=Llama 8b v3.12025.02 | 44.4 | |
| DegMat NLI Score Entail.Model=Llama 8b v3.12025.02 | 42 | |
| SATRMD+MSPModel=Llama 8b v3.12025.02 | 41.9 | |
| Semantic EntropyModel=Llama 8b v3.12025.02 | 41.6 | |
| Eccentricity NLI Score Entail.Model=Llama 8b v3.12025.02 | 40.9 | |
| Lexical Similarity ROUGE-LModel=Llama 8b v3.12025.02 | 40.3 | |
| EigenScoreModel=Llama 8b v3.12025.02 | 40.2 | |
| FactoscopeModel=Llama 8b v3.12025.02 | 24.2 | |
| SAPLMAModel=Llama 8b v3.12025.02 | 8.2 |