Question Answering on KairosBench Dataset B
72.5Fact Adherence ScoreKairosVL (7B)
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| KairosVL (7B)Open-Source=true2026.02 | 72.5 | 76.2 | 71.8 | 56.4 | |
| Qwen2.5VL-72BOpen-Source=true2026.02 | 71.2 | 42.4 | 54.4 | 64.7 | |
| KairosVL-SFT (7B)2026.02 | 66.8 | 73.6 | 64.9 | 51.6 | |
| Gemini-2.5FlashOpen-Source=false2026.02 | 65.6 | 55 | 52.9 | 53.1 | |
| Qwen2.5VL-32BOpen-Source=true2026.02 | 61.9 | 51.9 | 49.6 | 46.2 | |
| GLM4V-PlusOpen-Source=false2026.02 | 60.8 | 54 | 50 | 50.5 | |
| GPT-4oOpen-Source=false2026.02 | 57.1 | 63.3 | 58.3 | 56.9 | |
| Qwen2.5VL-7BOpen-Source=true2026.02 | 47.6 | 50 | 36.9 | 42.6 | |
| InternVL3-8BOpen-Source=true2026.02 | 41.3 | 53.6 | 40.4 | 42.7 |