Next Query Prediction on NQP-Bench Priv
44.22Judge ScoreOnePred
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OnePredModel Regime=Gemini-3.1-Pro, History Context Strategy=Ours2026.05 | 44.22 | 44.85 | |
| OnePredModel Regime=RL-trained Qwen, History Context Strategy=Ours2026.05 | 42.15 | 42.72 | |
| Gemini-3.1-ProModel Regime=Gemini-3.1-Pro, History Context Strategy=Current-turn2026.05 | 41.18 | 42.1 | |
| Gemini-3.1-ProModel Regime=Gemini-3.1-Pro, History Context Strategy=Full-history2026.05 | 40.81 | 41.15 | |
| RL-trained QwenModel Regime=RL-trained Qwen, History Context Strategy=Full-history2026.05 | 40.79 | 40.64 | |
| RL-trained QwenModel Regime=RL-trained Qwen, History Context Strategy=Current-turn2026.05 | 38.17 | 38.63 | |
| OnePredModel Regime=Base Qwen, History Context Strategy=Ours2026.05 | 33.46 | 33.18 | |
| Base QwenModel Regime=Base Qwen, History Context Strategy=Full-history2026.05 | 31.37 | 31.85 | |
| Base QwenModel Regime=Base Qwen, History Context Strategy=Current-turn2026.05 | 28.92 | 30.07 |