Chart Reasoning on CharXiv Rea
46.4AccuracyQwen2.5-VL-7B-Instruct + GRPO + KAWHI
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen2.5-VL-7B-Instruct + GRPO + KAWHITraining Method=GRPO + KAWHI2026.03 | 46.4 | |
| Qwen2.5-VL-7B-Instruct + GRPOTraining Method=GRPO2026.03 | 44.3 | |
| Qwen2.5-VL-7B-InstructModel Size=7B2026.03 | 42.1 |