Anat Veikherman

Papers that credit Anat Veikherman in their source bylines. Matching names may refer to different researchers.

1 results

1. PaperComputer Vision

   Many tested object-attribute errors show detectable visual evidence

   VELens probes visual evidence in two vision-language models and estimates that many tested object-attribute errors occur despite detected evidence.

   Added to Chiosa Oct 22, 2026