AutoData: EMNLP paper uses agentic search to automate pre-training data selection

ChengleiSi · x · 2026-09-18

Weco AI's new EMNLP paper, AutoData: Agentic Search for Pre-training Data Selection, asks whether autoresearch agents can find better training data, not just better training code. The paper frames pre-training data selection itself as a search problem solvable by agentic search methods.

Related event: AutoData: Agentic Search for Better Pre-training Data Selection(2 posts)→

Original post →

More from coding & agent

coding & agent channel →