Conversational Search on ClariQ
0.3625ROUGE-LRAC_SFT
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| RAC_SFTTraining=Supervised Fine-Tuning2026.01 | 0.3625 | 0.1488 | 0.3401 | 0.8952 | 0.5132 | 0.5315 | |
| RAC_DPOTraining=Direct Preference Optimization2026.01 | 0.3552 | 0.1486 | 0.3384 | 0.8939 | 0.5241 | 0.5577 | |
| TB2026.01 | 0.355 | 0.0028 | 0.2426 | 0.8765 | — | — | |
| QF-Cond2026.01 | 0.337 | 0.022 | 0.3756 | 0.8908 | — | — | |
| Q-Cond2026.01 | 0.2868 | 0.1119 | 0.2547 | 0.8816 | — | — | |
| QP-ZeroshotMode=Zero-shot2026.01 | 0.2603 | 0.0499 | 0.3181 | 0.8659 | — | — | |
| AT-CoTReasoning=Chain-of-Thought2026.01 | 0.1863 | 0.0349 | 0.2119 | 0.8474 | — | — |