List your company

Feedback Generation Evaluation on Expert Evaluation Feedback Quality

4.01AccuracyPsyScore
2.83483.13993.4453.7501Jun 18, 2026

Evaluation results

MethodLinks
PsyScoreModel=PsyScore (Full)2026.064.014.134.214.174.25
PsyScore-No-IRTAblation=No IRT layer2026.063.654.033.483.843.77
Qwen-3Model=Qwen-32026.063.132.912.912.73.11
GPT-4oModel=GPT-4o2026.063.12.882.792.553.15
Llama-4Model=Llama-42026.062.882.422.662.342.62