Boolean Classification of Social impact on Criminal analysis dataset
91.6AccuracyQwen-max
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen-maxModel Category=Full-sized Large Model2025.12 | 91.6 | 87.43 | 90.65 | |
| Gemini2.5-proModel Category=Full-sized Large Model2025.12 | 90.27 | 87.42 | 88.15 | |
| Deepseek v3Model Category=Full-sized Large Model2025.12 | 88.86 | 67.35 | 68.18 | |
| Claude-4Model Category=Full-sized Large Model2025.12 | 79.64 | 76.18 | 76.93 | |
| Gpt-5Model Category=Full-sized Large Model2025.12 | 78.35 | 74.82 | 75.24 | |
| Qwen2.5-7B-loraModel Category=Small Model, Adaptation=LoRA2025.12 | 71.15 | 67.35 | 68.18 |