Aggregate Data Analysis Performance on DABench, TableBench, and BIRD Combined
71.16Pass@1DATAMIND
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DATAMINDBackbone=Qwen-2.5-Coder-14B2025.09 | 71.16 | 80.25 | |
| ReActBackbone=DeepSeek-V3.12025.09 | 70.58 | 79.76 | |
| ReActBackbone=GPT-52025.09 | 69.44 | 76.26 | |
| ReActBackbone=o4-mini2025.09 | 69.06 | 77.93 | |
| DATAMINDBackbone=Qwen-2.5-Coder-7B2025.09 | 68.1 | 79.07 | |
| ReActBackbone=DeepSeek-R12025.09 | 67.83 | 77.75 | |
| ReActBackbone=Qwen-2.5-72B2025.09 | 67.02 | 77.36 | |
| ReActBackbone=GPT-4o2025.09 | 63.85 | 73.96 | |
| ReActBackbone=Llama-3.3-70B2025.09 | 61.45 | 73.03 | |
| ReActBackbone=QwQ-32B2025.09 | 59.42 | 74.87 | |
| ReActBackbone=Qwen-2.5-Coder-32B2025.09 | 58.49 | 71.25 | |
| ReActBackbone=Qwen-2.5-Coder-14B2025.09 | 56.64 | 71.05 | |
| OmniSQLBackbone=Qwen-2.5-Coder-14B2025.09 | 42.41 | 53.13 | |
| SQL-R1Backbone=Qwen-2.5-Coder-14B2025.09 | 42.16 | 52.7 | |
| OmniSQLBackbone=Qwen-2.5-Coder-7B2025.09 | 41.17 | 50.91 | |
| SQL-R1Backbone=Qwen-2.5-Coder-7B2025.09 | 40.83 | 50.5 | |
| Table-R1Backbone=Qwen-2.5-Coder-7B2025.09 | 36.53 | 52.03 | |
| Table-R1Backbone=Qwen-2.5-Coder-14B2025.09 | 35.84 | 50.79 | |
| TableLLMBackbone=Qwen-2.5-Coder-14B2025.09 | 35.23 | 59.89 | |
| TableLLMBackbone=Qwen-2.5-Coder-7B2025.09 | 29.9 | 53.03 | |
| ReActBackbone=Qwen-2.5-Coder-7B2025.09 | 11.26 | 27.58 |