Fact Verification on PubHealthTab OOD (test)
77.14AccuracyQwen3-VL-8B-DISCO
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-VL-8B-DISCOReasoning Strategy=Table-GLS, # TIR=02026.02 | 77.14 | |
| MiniCPM-V-2.6 8BReasoning Strategy=HIPPO, # TIR=55.2K2026.02 | 73.32 | |
| Qwen3-VL-8B-DISCOReasoning Strategy=DA, # TIR=02026.02 | 63.34 | |
| Gemma3n-E4B-DISCOReasoning Strategy=Table-GLS, # TIR=02026.02 | 62.41 | |
| Gemma3n-E4BReasoning Strategy=DA, # TIR=02026.02 | 56.08 | |
| Gemma3n-E4B-DISCOReasoning Strategy=DA, # TIR=02026.02 | 52.47 | |
| Table-LLaVA 7BReasoning Strategy=SFT, # TIR=82K2026.02 | 51.03 | |
| GPT-4o-miniReasoning Strategy=DA, # TIR=02026.02 | 48.61 | |
| Table-LLaVA 13BReasoning Strategy=SFT, # TIR=82K2026.02 | 48.46 | |
| Qwen3-VL-8BReasoning Strategy=DA, # TIR=02026.02 | 44.18 |