DCA-Bench: A Benchmark for Dataset Curation Agents

Posted: May 1st 2025
TL;DR

A benchmark exploring the performance of LLM Agents on detecting issues in datasets hosted on popular platforms.

DCA-Bench: A Benchmark for Dataset Curation Agents

It’s a pity that I won’t be able to attend KDD 2025 due to visa issues. My amazing mentor Jiaqi W. Ma and co-author Xingjian Zhang will be there on our behalf.

08/05/2025 Update:

  • I’m happy to hear that DCA-Bench received a total of six questions during the oral session. Surprisingly, the session chair showed great interest in our work—haha!
Last Updated on Aug 10th 2025 Powered by greatest-gatsby-academic-template.