Visual Math Reasoning on Geo3K
51.58AccuracyQwen2.5-VL-7B-Instruct + VEPO
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen2.5-VL-7B-Instruct + VEPOModel Access=Open-source, Model Scale=7B, Learning Strategy=VEPO2026.06 | 51.58 | |
| VPPO-7BModel Access=Open-source, Model Scale=7B, Learning Strategy=Visual-focused RL Fine-tuning2026.06 | 50.42 | |
| PAPO-DAPO-7BModel Access=Open-source, Model Scale=7B, Learning Strategy=Visual-focused RL Fine-tuning2026.06 | 50.08 | |
| Qwen2.5-VL-7B-Instruct + GRPOModel Access=Open-source, Model Scale=7B, Learning Strategy=GRPO2026.06 | 49.42 | |
| NoisyRollout-7BModel Access=Open-source, Model Scale=7B, Learning Strategy=Visual-focused RL Fine-tuning2026.06 | 48.42 | |
| Qwen2.5-VL-7B-Instruct + GRPO (Top 20% high entropy)Model Access=Open-source, Model Scale=7B, Learning Strategy=GRPO, Entropy Selection Strategy=Top 20% high entropy2026.06 | 47.59 | |
| Qwen2.5-VL-7B-Instruct + GRPO (Top 40% high entropy)Model Access=Open-source, Model Scale=7B, Learning Strategy=GRPO, Entropy Selection Strategy=Top 40% high entropy2026.06 | 46.76 | |
| R1-ShareVLModel Access=Open-source, Model Scale=7B, Learning Strategy=Visual-focused RL Fine-tuning2026.06 | 46.59 | |
| Qwen2.5-VL-7B-InstructModel Access=Open-source, Model Scale=7B2026.06 | 35.94 |