Therapeutic response evaluation on Mind Companion Expert Study Subset (10 evaluation points)
6.6Understanding ScoreGPT 5.2
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GPT 5.2Judge=LLM Judge2026.06 | 6.6 | 6.5 | 6.5 | 7 | 6.65 | 100 | |
| GPT 5.2Judge=Experts2026.06 | 5.45 | 5.23 | 5.41 | 5.1 | 5.3 | 90 | |
| Human TherapistJudge=LLM Judge2026.06 | 4.2 | 4.4 | 3.9 | 5.4 | 4.47 | 40 | |
| Human TherapistJudge=Experts2026.06 | 3.91 | 3.49 | 3.28 | 3.38 | 3.52 | 38 |