List your company

Table Question Answering on WikiTQ (Accuracy, Average Improvement, Cohen's d)

77.2AccuracyDataFactory
30.50442.62754.7566.873Mar 10, 2026

Evaluation results

MethodLinks
DataFactoryBackbone LLM=Gemini 2.5 Flash2026.0377.272.88.71
TabSQLifyBackbone LLM=GPT-4o mini2026.037469.65.24
DataFactoryBackbone LLM=Deepseek V32026.0373.572.88.71
StructGPTBackbone LLM=Gemini 2.5 Flash2026.0370.962.18.03
ReAcTableBackbone LLM=Gemini 2.5 Flash2026.0370.461.74.61
MACTBackbone LLM=Deepseek V32026.0368.666.14.8
AutoPrepBackbone LLM=Gemini 2.5 Flash2026.0368.267.22.57
TabSQLifyBackbone LLM=Deepseek V32026.036869.65.24
DataFactoryBackbone LLM=GPT-4o mini2026.0367.772.88.71
AutoPrepBackbone LLM=Deepseek V32026.0367.567.22.57
MACTBackbone LLM=Gemini 2.5 Flash2026.0367.366.14.8
TabSQLifyBackbone LLM=Gemini 2.5 Flash2026.0366.769.65.24
AutoPrepBackbone LLM=GPT-4o mini2026.0365.967.22.57
StructGPTBackbone LLM=Deepseek V32026.0365.562.18.03
MACTBackbone LLM=GPT-4o mini2026.0362.366.14.8
TableCoTBackbone LLM=Deepseek V32026.0360.64513.65
ReAcTableBackbone LLM=GPT-4o mini2026.0357.861.74.61
End-to-EndBackbone LLM=Gemini 2.5 Flash2026.0357.154.616.59
ReAcTableBackbone LLM=Deepseek V32026.0356.861.74.61
TAPEX-large2026.0356.256.29.76
End-to-EndBackbone LLM=Deepseek V32026.0355.954.616.59
Few-ShotBackbone LLM=Deepseek V32026.0355.649.613.94
TAPAS-large2026.0350.750.77.89
End-to-EndBackbone LLM=GPT-4o mini2026.0350.754.616.59
Few-ShotBackbone LLM=Gemini 2.5 Flash2026.0350.649.613.94
StructGPTBackbone LLM=GPT-4o mini2026.0349.862.18.03
Few-ShotBackbone LLM=GPT-4o mini2026.0342.749.613.94
TableCoTBackbone LLM=Gemini 2.5 Flash2026.0342.24513.65
TableCoTBackbone LLM=GPT-4o mini2026.0332.34513.65