Table Question Answering on VWTQ (test)
67.5AccuracyGPT-4o
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-4oModel Scale=Large Scale2026.06 | 67.5 | |
| VTool-R1 (GRPO)Model Scale=Large Scale, Backbone=Qwen-2.5VL-7B2026.06 | 52.6 | |
| IAPOModel Scale=Small Scale, Backbone=Qwen-2.5VL-3B2026.06 | 46.7 | |
| VTool-R1 (GRPO)Model Scale=Small Scale, Backbone=Qwen-2.5VL-3B2026.06 | 46.6 | |
| ToRLModel Scale=Small Scale, Backbone=Qwen-2.5VL-3B2026.06 | 44.4 | |
| TRMModel Scale=Small Scale, Backbone=Qwen-2.5VL-3B2026.06 | 44.3 | |
| Qwen-2.5VL-7BModel Scale=Large Scale2026.06 | 37.5 | |
| Qwen-2.5VL-3BModel Scale=Small Scale2026.06 | 14.2 |