Response reliability estimation on Multiturn
0.831AUROCGLU
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GLUModel=Fanar2026.06 | 0.831 | 0.687 | |
| RAUQModel=Fanar2026.06 | 0.804 | 0.654 | |
| GLUModel=Qwen2026.06 | 0.739 | 0.407 | |
| RAUQModel=Gemma2026.06 | 0.734 | 0.506 | |
| GLUModel=Gemma2026.06 | 0.701 | 0.423 | |
| RAUQModel=Qwen2026.06 | 0.644 | 0.189 | |
| LogTokUModel=Qwen2026.06 | 0.523 | 0.008 | |
| LogTokUModel=Fanar2026.06 | 0.463 | 0.166 | |
| LogTokUModel=Gemma2026.06 | 0.457 | 0.274 |