Talk:Differentiable imaging

Feedback from New Page Review process

I left the following feedback for the creator/future reviewers while reviewing this article: I am highly dubious about this article for three reasons: 1) it is AI, 2) it is 95% the work of one group, and 3) it does a poor job of history. Modelling microscopes (particularly electron) digitally goes back to the 1980's or earlier. I see no mention if anything but one groups work.

Ldm1954 (talk) 13:25, 23 September 2025 (UTC)

I do agree with 3) and especially 2), but I'm less certain about 1). (I do think the prose could be greatly improved, of course.) I did see enough other work when I was reviewing that I wasn't worried about other research not existing and wasn't a blocker for me accepting. (Research for MIRT/MIRG which IIRC has some involvement with differentiable imaging.) That being said, it is all outside of anything I was already familiar with, so I'll defer to you on anything here. :) Perryprog (talk) 13:57, 23 September 2025 (UTC)

Thank you for raising these issues. I agree the article currently relies too heavily on one group’s publications, and I will work on adding broader historical context (e.g., electron microscope multislice modeling, tomography methods) and citing independent sources. I’ll also revise the text to improve clarity and neutrality, and make sure the focus on modeling uncertainties and joint inverse solving is more clearly distinguished from earlier digital modeling work. 147.8.183.189 (talk) 00:05, 24 September 2025 (UTC)