Question Answering on ARC Challenge (Action Rate and Risk Assessment)
0Pathwise Violations (PathV)CSA
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CSAN (EVAL size)=938, Err (Base verifier-fail rate)=10.0%, alpha* (Target risk level)=0.102026.05 | 0 | 56.2 | 7.8 | |
| NEX-ConfN (EVAL size)=938, Err (Base verifier-fail rate)=10.0%, alpha* (Target risk level)=0.102026.05 | 20 | — | — | |
| 5 online baselinesN (EVAL size)=938, Err (Base verifier-fail rate)=10.0%, alpha* (Target risk level)=0.102026.05 | 100 | — | — |