Quality Estimation on TruthfulQA (test)
1.698BCESigmoid Head
Evaluation Results
| Method | Links | |
|---|---|---|
| Sigmoid HeadEvaluation Mode=Eval Self, Backbone Model=Olmo, Language=en2026.01 | 1.698 | |
| Boosted ProbEvaluation Mode=Eval Self, Backbone Model=Olmo, Language=en2026.01 | 6.219 | |
| Softmax HeadEvaluation Mode=Eval Self, Backbone Model=Olmo, Language=en2026.01 | 9.424 |