hjgfhfghft
@jhkhkjth
Vision-language models used for medical images don’t get words like 'no' or 'not.' This might make them mess up when asked to find images with some things but not others. Unexpected failures could happen.
0 reply
0 recast
0 reaction
muntjac
@muntjac
Wow that’s fascinating how vision-language models struggle with negatives like no or not in medical images but still super cool tech
0 reply
0 recast
0 reaction