Uncertainty Quantification on Multiturn (PRR)
0.777PRRP(true)
Evaluation Results
| Method | Links | |
|---|---|---|
| P(true)Model=Qwen2026.06 | 0.777 | |
| P(true)Model=Gemma2026.06 | 0.723 | |
| GLU-AUModel=Fanar2026.06 | 0.691 | |
| GLUModel=Fanar2026.06 | 0.687 | |
| RAUQModel=Fanar2026.06 | 0.654 | |
| LogProbModel=Fanar2026.06 | 0.622 | |
| LogProbModel=Gemma2026.06 | 0.524 | |
| RAUQModel=Gemma2026.06 | 0.506 | |
| GLU-DKModel=Gemma2026.06 | 0.451 | |
| GLU-S¯Model=Qwen2026.06 | 0.444 | |
| GLUModel=Gemma2026.06 | 0.423 | |
| GLUModel=Qwen2026.06 | 0.407 | |
| P(true)Model=Fanar2026.06 | 0.252 | |
| RAUQModel=Qwen2026.06 | 0.189 | |
| LogProbModel=Qwen2026.06 | 0.131 | |
| LogTokUModel=Qwen2026.06 | 0.008 | |
| LogTokUModel=Fanar2026.06 | -0.166 | |
| LogTokUModel=Gemma2026.06 | -0.274 |