#Geospatialml

We reviewed 22 ML conference submissions this summer. Fifteen had fabricated citations, hallucinated authors, or clear LLM slop, so we complain about that a bit, and release the paper references audit we now run.