Selective Prediction on CoQA
80.6PRRSP×ExpectedRank
Evaluation Results
| Method | Links | |
|---|---|---|
| SP×ExpectedRankGenerator=GPT, Optimization=Evolved specifically for selective prediction2026.04 | 80.6 | |
| Evo-Gpt-9Generator=GPT2026.04 | 80.5 | |
| Evo-Gpt-7Generator=GPT2026.04 | 80.5 | |
| SequenceProbability2026.04 | 80.5 | |
| RAUQ2026.04 | 80.2 | |
| SP×MeanEntropyGenerator=GPT, Optimization=Evolved specifically for selective prediction2026.04 | 79.8 | |
| Evo-Anth-22Generator=Claude2026.04 | 79.4 | |
| Evo-Anth-15Generator=Claude2026.04 | 79.1 | |
| Evo-Anth-30Generator=Claude2026.04 | 79 | |
| Evo-Anth-1Generator=Claude2026.04 | 77.5 | |
| Evo-Anth-8Generator=Claude2026.04 | 76.9 | |
| MeanTokenEntropy2026.04 | 76.7 | |
| Perplexity2026.04 | 76.3 | |
| Evo-Gpt-13Generator=GPT2026.04 | 75.8 | |
| Evo-Gpt-5Generator=GPT2026.04 | 73.3 | |
| Evo-Gpt-1Generator=GPT2026.04 | 73.1 | |
| AttentionScore2026.04 | 72.8 | |
| Evo-Gpt-11Generator=GPT2026.04 | 72.2 | |
| Evo-Gpt-15Generator=GPT2026.04 | 71.6 | |
| Evo-Gpt-3Generator=GPT2026.04 | 68.5 |