TableMage-Analysis

Benchmarking details and scripts for ChatDA from the paper "Tool-wielding language model-based agent offers conversational exploration of clinical tabular data".

Running DataAnalysisQA

Please see ./benchmarking/dataanalysisqa/README.md for details.

Running the machine learning benchmark

Please see ./benchmarking/ml/README.md for details.

Citation

@article{Yang2025-TableMage,
    author = {Yang, Andrew and Woo, Joshua and Zhang, Ryan and Mach, Alan and Ramkumar, Prem and Ma, Ying},
    title = {Tool-wielding language-model-based agent offers conversational exploration of clinical tabular data},
    elocation-id = {2025.12.01.25341392},
    year = {2025},
    doi = {10.64898/2025.12.01.25341392},
    publisher = {Cold Spring Harbor Laboratory Press},
    URL = {https://www.medrxiv.org/content/early/2025/12/02/2025.12.01.25341392},
    eprint = {https://www.medrxiv.org/content/early/2025/12/02/2025.12.01.25341392.full.pdf},
    journal = {medRxiv}
}