Conversation Performance on ReDial
5.1BLEU-2STARCRS
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| STARCRS2026.01 | 5.1 | 2.5 | 6.6 | 29.5 | 78.1 | 119.2 | 141.9 | |
| DCRS2026.01 | 4.8 | 2.4 | 6.3 | 28.5 | 77.9 | 117.3 | 138.6 | |
| UNICRS2026.01 | 4.5 | 2.1 | 5.8 | 28.5 | 43.3 | 74.8 | 100.3 | |
| MSCRS2026.01 | 4.3 | 1.7 | 5.8 | 28.7 | 77 | 111.2 | 131 | |
| DialogGPT2026.01 | 4.1 | 2.1 | 5.4 | 25.8 | 43.6 | 63.2 | 77.1 | |
| KBRD2026.01 | 3.8 | 1.8 | 4.7 | 23.7 | 7 | 28.8 | 48.8 | |
| GPT-22026.01 | 3.1 | 1.3 | 4.1 | 24.4 | 40.5 | 60.3 | 75.7 | |
| KGSF2026.01 | 3 | 1.2 | 3.9 | 24.4 | 6.1 | 27.8 | 51.5 | |
| BART2026.01 | 2.4 | 1.1 | 3.1 | 22.9 | 43.2 | 61.5 | 70.5 | |
| TREA2026.01 | 2.2 | 0.8 | 3.9 | 17.5 | 24.2 | 61.5 | 117.6 | |
| VRICR2026.01 | 2.1 | 0.8 | 2.7 | 13.7 | 10.7 | 28.6 | 47.1 | |
| ReDial2026.01 | 0.4 | 0.1 | 2.1 | 18.7 | 5.8 | 20.4 | 44.2 |