Question Answering on MuSR n=756 (test)
98.8Activation ScoreLearned Stopper
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Learned Stopperbeta=0.15, T=42026.06 | 98.8 | 63.5 | 36.5 | |
| Consensusbeta=0.15, T=42026.06 | 91.9 | 65.7 | 34.3 | |
| Oraclebeta=0.15, T=42026.06 | 68.9 | 100 | 0 | |
| Selective Predictionbeta=0.15, T=42026.06 | 43.4 | 67.2 | 14.2 | |
| CRCbeta=0.15, T=42026.06 | 43.3 | 67.2 | 14.2 | |
| Confidence Thresholdbeta=0.15, T=42026.06 | 32.9 | 66 | 34 | |
| Calibrated Learnedbeta=0.15, T=42026.06 | 18.9 | 46.8 | 6 | |
| Isotonic Confidencebeta=0.15, T=42026.06 | 7.6 | 30 | 3 | |
| Adaptive Local k-NN Certification (Ours)beta=0.15, T=42026.06 | 0 | — | 0 | |
| kNN no biasbeta=0.15, T=42026.06 | 0 | — | 0 | |
| Final-roundbeta=0.15, T=42026.06 | 0 | — | 0 |