Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Visual Math Reasoning on Geo3K
Loading...
51.58
Accuracy
Qwen2.5-VL-7B-Instruct + VEPO
35.3144
39.5372
43.76
47.9828
Jun 2, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Qwen2.5-VL-7B-Instruct + VEPO
Model Access=Open-sour...
2026.06
51.58
VPPO-7B
Model Access=Open-sour...
2026.06
50.42
PAPO-DAPO-7B
Model Access=Open-sour...
2026.06
50.08
Qwen2.5-VL-7B-Instruct + GRPO
Model Access=Open-sour...
2026.06
49.42
NoisyRollout-7B
Model Access=Open-sour...
2026.06
48.42
Qwen2.5-VL-7B-Instruct + GRPO (Top 20% high entropy)
Model Access=Open-sour...
2026.06
47.59
Qwen2.5-VL-7B-Instruct + GRPO (Top 40% high entropy)
Model Access=Open-sour...
2026.06
46.76
R1-ShareVL
Model Access=Open-sour...
2026.06
46.59
Qwen2.5-VL-7B-Instruct
Model Access=Open-sour...
2026.06
35.94
Feedback
Search any
task
Search any
task