Google's TabFM-Auto: LLM agent evolves data pipelines, lifting TabFM by 228 Elo and topping MLE-Bench

google · hf · 2026-09-30

Google introduces TabFM-Auto, pairing the TabFM tabular foundation model with an LLM agent that iteratively evolves data cleaning, feature engineering, and post-processing guided by metadata and validation feedback. Its five configurations sweep the top five on all 51 TabArena datasets, raising TabFM from 1785 to 2013 Elo; pipelines transfer to other frozen tabular models (+69 to +143 Elo). It ranks first among MLE agents on MLE-Bench's 8 tabular competitions.

Original post →

More from coding & agent

coding & agent channel →