ESConv
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Emotional Support Conversation | ESConv (test) | BLEU-222.85 | 50 | |
| Human Evaluation | ESConv | Win Rate70 | 36 | |
| Emotional Support Conversation | ESConv | BERT-SCORE85.89 | 24 | |
| Emotional Support Dialogue | ESConv | Success Rate (SR)52.5 | 16 | |
| Rationale-based retrieval | ESConv (test) | R@134.4 | 15 | |
| Emotional Support Conversation | ESConv | PPL13.1 | 15 | |
| Response Generation | ESConv (test) | Fluency3.9 | 10 | |
| Proactive dialogue | ESConv | Success Rate73.85 | 10 | |
| Emotional Support Conversation Response Generation | ESConv (test) | BLEU-121.83 | 10 | |
| Emotional Support Dialogue | ESConv (test) | Engagingness-12.1 | 7 | |
| Open-ended conversation | ESConv (out-of-domain) | ROUGE-L15.62 | 6 | |
| Strategy Prediction | ESConv | Macro F110.52 | 6 | |
| Emotional Support Conversation | ESConv | Fluency187 | 6 | |
| Emotional Support Conversation | ESConv | Perplexity16.16 | 6 | |
| Emotional Support Conversation Generation | ESConv (test) | Accuracy34.51 | 5 | |
| Emotional Support Dialogue | ESConv | FES55.1 | 4 | |
| Emotional Support Conversation | ESConv (test) | Identification Score2.18 | 3 | |
| Emotional Support Conversation | ESConv | Empathy99.73 | 3 | |
| Emotional Support Dialogue | ESConv | Win Rate vs PRINC.58 | 2 | |
| Response Generation | ESConv | BLEU0.005 | 2 |
Showing 20 of 20 rows