Science Question Answering on SciQ (Score, Delta (Δ), Avg. Rank)
98.53ScoreQwen2.5(14B)T
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen2.5(14B)TEvaluation Protocol=Generation-based2026.07 | 98.53 | — | — | |
| RG-OPDEvaluation Protocol=Generation-based2026.07 | 97.8 | 1.07 | 2 | |
| Reverse KLEvaluation Protocol=Generation-based2026.07 | 97.37 | 0.64 | 3 | |
| TSD-KDEvaluation Protocol=Generation-based2026.07 | 97.2 | 0.47 | 2.6 | |
| Qwen2.5(1.5B)SEvaluation Protocol=Generation-based2026.07 | 96.73 | 0 | 2.4 | |
| TSD-KDEvaluation Protocol=Likelihood-based2026.07 | 93.6 | 4.1 | 2.6 | |
| Qwen2.5(1.5B)SEvaluation Protocol=Likelihood-based2026.07 | 89.5 | 0 | 2.4 | |
| Qwen2.5(14B)TEvaluation Protocol=Likelihood-based2026.07 | 86.8 | — | — | |
| Reverse KLEvaluation Protocol=Likelihood-based2026.07 | 80 | -9.5 | 3 | |
| RG-OPDEvaluation Protocol=Likelihood-based2026.07 | 79 | -10.5 | 2 |