TMLR

Navigating peer review timelines: what to do when a third review is delayed

A month of silence after two reviews would test anyone's patience, especially when you've already acted on the feedback.

4 min readMachine Learning

The wait for a third review in TMLR is a familiar kind of limbo, and the quiet anxiety that comes with it is captured. Two reviews arrived quickly, the revisions are done, and yet the process has stalled for a month with no response from the Action Editor. This is not a complaint about the system, and the author is careful to say so. It is a practical question about a very real bottleneck: when do you keep waiting, and when do you start planning around silence?

This is the unglamorous side of peer review that no one prepares you for. The author has done everything right, responded to feedback, followed the journal's guidance to wait for all three reviews before posting updates, and even reached out to the Action Editor once. That the editor has not replied is not a signal of failure, but it is a reminder that the process is human, and humans get overloaded. We have seen this dynamic play out in other corners of the research world, whether it is the pressure of ICLR Submissions Exposed: Addressing Data Privacy Concerns in AI Research or the practical realities of Exploring Real-World Computer Vision: Deployments, Edge Models, and Current Challenges. The pattern is consistent: the system works, but it works slowly, and that slowness is not always within your control.

What should this author do? The honest answer is to keep waiting, but not passively. A month of silence after a direct message is enough time to send a polite follow-up. That is not nagging; it is a professional courtesy to the editor, who may have lost track of the thread. In the meantime, the author should treat the two reviews as actionable. If the revisions are solid and the third review is delayed, there is no harm in preparing a response letter that addresses both reviewers in detail. When the third review does arrive, and it will, the author will be ready to post immediately. That is the practical move: use the waiting time to build momentum, not to sit and refresh OpenReview.

The deeper takeaway here is about managing expectations in a system that prizes rigor over speed. TMLR's approach, waiting for multiple reviews before allowing authors to respond, is deliberate. It prevents a fragmented conversation and keeps the focus on the final version. But it also means that authors surrender some control over the timeline. The instinct to ask, "Is this normal?" is the right one. is the right one. The answer is yes, and the next step is to accept that the process is working, even when it does not feel like it. The specific thing to watch is the editor's response to the follow-up. If that goes unanswered for another month, it may be worth a gentle nudge to the program chairs. That is not a threat; it is a practical escalation that respects everyone's time. For now, the author should trust the process, keep the revisions polished, and remember that a slow review is not a rejection. It is just a pause, and pauses are part of the rhythm.

From Machine Learning

I have a submission under review in TMLR. Less than a month after submission, I have already received 2 reviews. However, a month has passed since those two reviews, and I still haven't received the third. Is this normal?

I have already made the suggested changes, but the journal recommends updating the manuscript in OpenReview only after receiving three reviews. So I also haven't posted any answers yet. I messaged the Action Editor to ask whether I would receive another review, but I sent it a month ago, and he hasn't answered.

Read the original at Machine Learning