What Changed Since the Last Review? Comparing Manuscript Revisions
When an author resubmits, ReviewerZero compares the revision against the previous review and shows exactly what is resolved, what is still open, and what is new, so a second look takes minutes instead of a full re-read.
A revised manuscript arrives with a cover letter that says every reviewer comment was addressed. In principle that is good news, but it is also a claim, and checking it means reading the whole paper again to find what actually moved. ReviewerZero now does that comparison for you.
Our newest feature lines the revision up against the previous review and sorts every earlier finding into three buckets. What is resolved, what is still open, and what is new.
Why re-reviewing a revision is hard
A first review starts from a blank page while a second review does not. Editors and reviewers are interested now in "what changed since last time, and did it get better or worse." Answering that by hand is slow and easy to get wrong.
A revision can also quietly make things worse. Authors fix the issues you raised and, along the way, add a new citation that turns out to be fabricated, introduce a statistic that does not add up, or paste in a paragraph that reads as machine generated.
How ReviewerZero compares versions
Doing a diff between versions sounds simple, but it is actually quite complex. At ReviewerZero, we do not diff the text. We match findings with a bit of engineering.
Every issue from the previous review is given a stable identity based on what it is about (the cited work behind a reference, the specific statistic behind a p-value check, the paragraph behind an AI-text flag, and so on) and not where it sits on the page. When we analyze the new version, each finding is matched to its counterpart in the old one by that identity. Anything flagged before and gone now is resolved, anything flagged in both is still open, and anything flagged only in the new version is new. It is a bit more complicated than that, but you get the idea.
Because the match is on content rather than position, an issue that moved to a new section or came back slightly reworded is still recognized as the same issue, not mistaken for a fresh one.
Decisions you already made carry over
If you dismissed a finding in the first round because it was a known limitation or for some other reason, you should not have to dismiss it again. When the same finding is still present in the revision, unchanged, your earlier decision comes with it, marked as carried over so you know it was inherited rather than re-judged.
What you see
At the top of a revised submission's report, the revision summary leads with the verdict and the counts, followed by every changed finding grouped into new, still open, and resolved. Every finding sits in the same evenly-weighted list. A newly introduced problem, even a serious one like a retracted reference, appears alongside everything else rather than being singled out or pushed to the top.
Try it
Revisions work the way the rest of ReviewerZero does. Submit a revision from a finished report, or connect your submission system so resubmissions are linked automatically, and the comparison is ready as soon as the analysis finishes.
Book a demo to run reference, statistics, figure, and revision checks on your own papers.
