Emotional Support Dialogue on SAGE
72.1Average ScoreESC-Skills
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ESC-SkillsBackbone=Qwen3.6-Plus, Skill Augmentation=ESC-Skills2026.05 | 72.1 | 31 | 12 | |
| Qwen3.6-PlusBackbone=Qwen3.6-Plus, Skill Augmentation=None2026.05 | 66.4 | 13 | 14 | |
| ESC-SkillsBackbone=Claude-Sonnet-4.6, Skill Augmentation=ESC-Skills2026.05 | 63.6 | 11 | 21 | |
| ESC-SkillsBackbone=Claude-Opus-4.6, Skill Augmentation=ESC-Skills2026.05 | 61.8 | 19 | 21 | |
| Claude-Opus-4.6Backbone=Claude-Opus-4.6, Skill Augmentation=None2026.05 | 61.2 | 16 | 18 | |
| Claude-Sonnet-4.6Backbone=Claude-Sonnet-4.6, Skill Augmentation=None2026.05 | 58.2 | 9 | 23 | |
| ESC-SkillsBackbone=Gemini-3.1-Flash, Skill Augmentation=ESC-Skills2026.05 | 57.6 | 7 | 19 | |
| ESC-SkillsBackbone=GPT-5.4-0305-Global, Skill Augmentation=ESC-Skills2026.05 | 57.4 | 7 | 19 | |
| GPT-5.4-0305-GlobalBackbone=GPT-5.4-0305-Global, Skill Augmentation=None2026.05 | 56.9 | 6 | 21 | |
| Gemini-3.1-FlashBackbone=Gemini-3.1-Flash, Skill Augmentation=None2026.05 | 56.2 | 4 | 21 | |
| ESC-SkillsBackbone=Claude-Haiku-4.5, Skill Augmentation=ESC-Skills2026.05 | 42.3 | 8 | 43 | |
| Claude-Haiku-4.5Backbone=Claude-Haiku-4.5, Skill Augmentation=None2026.05 | 29.7 | 2 | 51 |