Practice Pathway Evaluation on SocialCoach Practice Sessions (Simulated) (test)
4.38Average EngagementSocialCoachQwen3-8B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SocialCoachQwen3-8BTuning Status=Finetuned, Backbone Model=Qwen3-8B2026.06 | 4.38 | 4.11 | 4.25 | 4.05 | |
| NonTuningGemini-3-ProTuning Status=NonTuning, Backbone Model=Gemini-3-Pro2026.06 | 3.95 | 3.88 | 4.08 | 3.78 | |
| NonTuningQwen3-235BTuning Status=NonTuning, Backbone Model=Qwen3-235B2026.06 | 3.92 | 3.68 | 3.96 | 3.67 | |
| NonTuningGPT-5Tuning Status=NonTuning, Backbone Model=GPT-52026.06 | 3.88 | 3.75 | 4.02 | 3.74 | |
| NonTuningDeepSeek-R1-0528Tuning Status=NonTuning, Backbone Model=DeepSeek-R1-05282026.06 | 3.76 | 3.72 | 3.92 | 3.68 | |
| NonTuningQwen3-8BTuning Status=NonTuning, Backbone Model=Qwen3-8B2026.06 | 3.61 | 3.44 | 3.78 | 3.42 | |
| ProfileQueryBackbone Model=ProfileQuery2026.06 | 3.45 | 3.52 | 3.61 | 3.4 |