Factuality Detection on EN Text IF
86.6ScoreYuvion VL-32B
Evaluation results
| Method | Links | |
|---|---|---|
| Yuvion VL-32BProtocol=Instruct2026.06 | 86.6 | |
| Yuvion VL-32BProtocol=Reasoning2026.06 | 86.3 | |
| Yuvion VL-8BProtocol=Reasoning2026.06 | 85.4 | |
| Qwen3.5-PlusModel Size=397B2026.06 | 82.6 | |
| K2.5Model Size=≈1T2026.06 | 82.2 | |
| Yuvion VL-8BProtocol=Instruct2026.06 | 82.1 | |
| GLM-5Model Size=≈700B2026.06 | 80.7 | |
| Qwen3-VL-32BProtocol=Thinking2026.06 | 80.1 | |
| GPT-5.4Model Size=≈1.5T2026.06 | 75.8 | |
| Qwen3-VL-32BProtocol=Instruct2026.06 | 70.8 | |
| Qwen3-VL-8BProtocol=Instruct2026.06 | 69.5 | |
| Qwen3-VL-8BProtocol=Thinking2026.06 | 69.5 |