Grounding & counting on Count Bench
93.7Pass@1Seed 1.5-VL
Evaluation Results
| Method | Links | |
|---|---|---|
| Seed 1.5-VLthinking=true, decoding=greedy2025.05 | 93.7 | |
| Qwen 2.5-VL 72Bthinking=false, decoding=greedy2025.05 | 93.6 | |
| Seed 1.5-VLthinking=false, decoding=greedy2025.05 | 93.5 | |
| Gemini 1.5 Prothinking=true, decoding=greedy2025.05 | 91 | |
| OpenAI o1thinking=true, decoding=greedy2025.05 | 86.6 | |
| Claude 3.7 Sonnetthinking=true, decoding=sampling2025.05 | 86.1 | |
| GPT-4othinking=false, decoding=greedy2025.05 | 85.7 |