Prosociality assessment on PSA (Prosociality) scale (test)
5.76Mean ScoreGPT-4
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GPT-4evaluation_protocol=Context-based evaluation, scale=1-7 Likert scale2026.01 | 5.76 | 0.87 | -0.06 | 1.02 | |
| GPT-3.5evaluation_protocol=Context-based evaluation2026.01 | 5.67 | 0.62 | -0.15 | 0.84 | |
| Llama3-70Bevaluation_protocol=Context-based evaluation2026.01 | 5.54 | 0.71 | -0.28 | 0.91 | |
| Mistral-7Bevaluation_protocol=Context-based evaluation2026.01 | 5.28 | 0.84 | -0.54 | 0.94 | |
| GPT-4oevaluation_protocol=Context-based evaluation2026.01 | 5.27 | 0.64 | -0.55 | 0.73 | |
| GPT-4-turboevaluation_protocol=Context-based evaluation2026.01 | 5.16 | 0.67 | -0.66 | 0.84 | |
| DeepSeek-R1evaluation_protocol=Context-based evaluation2026.01 | 4.94 | 1.43 | -0.88 | 1.16 | |
| Qwen3-14Bevaluation_protocol=Context-based evaluation2026.01 | 4.76 | 1.85 | -1.06 | 1.84 | |
| Amazon Titanevaluation_protocol=Context-based evaluation2026.01 | 4.65 | 1.46 | -1.17 | 1.52 |