Open Question Answering on OpenQA (Accuracy)
88.2AccuracyTrajectory
Evaluation Results
| Method | Links | |
|---|---|---|
| TrajectoryTrain Dataset=OpenQA, Model=Qwen2.5-14b2026.03 | 88.2 | |
| TrajectoryTrain Dataset=ARC-E, Model=Qwen2.5-14b2026.03 | 84.4 | |
| Linear ProbeTrain Dataset=OpenQA, Model=Qwen2.5-14b2026.03 | 80.31 | |
| Linear ProbeTrain Dataset=ARC-E, Model=Qwen2.5-14b2026.03 | 77.83 | |
| TrajectoryTrain Dataset=ARC-C, Model=Qwen2.5-14b2026.03 | 77.2 | |
| Linear ProbeTrain Dataset=ARC-C, Model=Qwen2.5-14b2026.03 | 75.48 | |
| Few-shot AccuracyModel=Qwen2.5-14b2026.03 | 62 | |
| Zero-shot AccuracyModel=Qwen2.5-14b2026.03 | 61.8 |