Dialogue Evaluation on ConsistentChat (test)
7.3G-E ScoreMDS
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MDSselection_source=Banking corpus, selection_budget=10K2026.04 | 7.3 | 30 | |
| Random Dataselection_source=Banking corpus, selection_budget=10K2026.04 | 7.22 | 29 | |
| Heuristicselection_source=Banking corpus, selection_budget=10K2026.04 | 7.22 | 27.8 | |
| DialScoreselection_source=Banking corpus, selection_budget=10K2026.04 | 7.2 | 29.1 | |
| Rethinkingselection_source=Banking corpus, selection_budget=10K2026.04 | 7.18 | 28.2 | |
| ZIPselection_source=Banking corpus, selection_budget=10K2026.04 | 7.16 | 28.5 | |
| CC-Scoreselection_source=Banking corpus, selection_budget=10K2026.04 | 7.16 | 28.5 | |
| All Dataselection_source=Banking corpus, selection_budget=Full2026.04 | 7.12 | 28.8 | |
| SuperFilteringselection_source=Banking corpus, selection_budget=10K2026.04 | 7.12 | 28.3 | |
| Backboneselection=none (base model)2026.04 | 6.7 | 22.2 |