'Lost to be found' started as a public intervention project, where commonly found notebooks were filled with abstract signs and symbols as part of my cognitive drawing practice, and then anonymously left in different parts of the world. Till date some 70+ notebooks have been lost in various public places, not bearing any marks of artist, so that drawings may be the only source of interpretation. There are no whereabouts of the original notebooks, if they remain, modified or are destroyed, and no stories of what happened when someone must have found them.

Some notebooks from this project were exhibited as part of Open Sessions 10 group exhibition 'Marginalia', at The Drawing Center in New York in 2017. (https://drawingcenter.org/exhibitions/marginalia-open-sessions-10)

In this series of works, I am curating pages from these scanned notebooks to be interpreted and reconstructed by ai algorithm trained on public ImageNet dataset using variational auto encoder and clip. On the left side is the original drawing, and on the right side is the image reconstructed by the ai model.