Automated CTRS evaluation on CACTUS, PsyDTCorpus, CCDCHAT (150 sampled cases)
5.2UnderstandingCCDCHAT
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| CCDCHATEvaluator=GPT-4o-mini, Sample size=150 cases2026.04 | 5.2 | 4.94 | 5.32 | 4.96 | 4.92 | 4.5 | |
| PsyDTCorpusEvaluator=GPT-4o-mini, Sample size=150 cases2026.04 | 5.06 | 4.8 | 5.3 | 4.92 | 4.92 | 4.24 | |
| CACTUSEvaluator=GPT-4o-mini, Sample size=150 cases2026.04 | 4.84 | 4.64 | 5.06 | 4.68 | 4.44 | 4.32 |