2% of ICML papers desk rejected because the authors used LLM in their reviews
1–10 of 172 posts
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#2Worth reading for the discussion of the LLM watermark technique alone.
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#3One thing to note.
They were quite conservative in their approach, so the only things that were rejected were from people who had agreed not to use an LLM and almost definitely did use an LLM (since they fed hidden watermarked instructions to the llm's).
This means the true number of people that used LLM's in their review (even in group A that had agreed not to) is likely higher.
Also worth noting, 10% of these authors used them in more than half of their reviews.
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#4I'm amazed that such a simple method of detection worked so flawlessly for so many people. This would not work for those who merely used LLMs to help pinpoint strengths and weaknesses in the paper; there are separate techniques to judge that. Instead, it only detects those who quite literally copied and pasted the LLM output as a review.
It's incredible how so many people thought it was fair that their paper should be assessed by human reviewers alone, and yet would not extend the same courtesy to others.
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#5To be clear, as the article says, these authors were offered a choice and agreed to be on the "no LLMs allowed" policy.
And detection was not done with some snake oil "AI detector" but by invisible prompt injection in the paper pdf, instructing LLMs to put TWO long phrases into the review. They then detected LLM use through checking if both phrases appear in the review.
This did not detect grammar checks and touchups of an independently written review. The phrases would only get included if the reviewer fed the pdf to the LLM in clear violation to their chosen policy.
> After a selection process, in which reviewers got to choose which policy they would like to operate under, they were assigned to either Policy A or Policy B. In the end, based on author demands and reviewer signups, the only reviewers who were assigned to Policy A (no LLMs) were those who explicitly selected “Policy A” or “I am okay with either [Policy] A or B.” To be clear, no reviewer who strongly preferred Policy B was assigned to Policy A.
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#6I'm amazed that such a simple method of detection worked so flawlessly for so many people. This would not work for those who merely used LLMs to help pinpoint strengths and weaknesses in the paper; there are separate techniques to judge that. Instead, it only detects those who quite literally copied and pasted the LLM output as a review. It's incredible how so many people thought it was fair that their paper should b…
Generally speaking people have worse impulse control than they believe they do. Once you give a tool that does most of the work for you, very very few people will actually be able to use that tool in truly enriching ways. The majority of people (even the smart ones) will weaken over time and take shortcuts.
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#7To be clear, as the article says, these authors were offered a choice and agreed to be on the "no LLMs allowed" policy. And detection was not done with some snake oil "AI detector" but by invisible prompt injection in the paper pdf, instructing LLMs to put TWO long phrases into the review. They then detected LLM use through checking if both phrases appear in the review. This did not detect grammar checks and touchups…
In that case, I hope these frauds have been banned for life.
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#8Another 30-40% just didn't get caught because the reviewers also used LLM in their "reviews"
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#9To be clear, as the article says, these authors were offered a choice and agreed to be on the "no LLMs allowed" policy. And detection was not done with some snake oil "AI detector" but by invisible prompt injection in the paper pdf, instructing LLMs to put TWO long phrases into the review. They then detected LLM use through checking if both phrases appear in the review. This did not detect grammar checks and touchups…
In that case, I hope these frauds have been banned for life.
I was thinking this too, but I don't believe this is the case, and I feel like it would not be a good idea either.
Most of these people are likely students; this should be a learning moment, but I don't think it is yet grounds for their entire academic career to be crippled by being unable to publish in a top-tier ML venue.
Re: 2% of ICML papers desk rejected because the authors used LLM in their reviews
#10I'm amazed that such a simple method of detection worked so flawlessly for so many people. This would not work for those who merely used LLMs to help pinpoint strengths and weaknesses in the paper; there are separate techniques to judge that. Instead, it only detects those who quite literally copied and pasted the LLM output as a review. It's incredible how so many people thought it was fair that their paper should b…
Generally speaking people have worse impulse control than they believe they do. Once you give a tool that does most of the work for you, very very few people will actually be able to use that tool in truly enriching ways. The majority of people (even the smart ones) will weaken over time and take shortcuts.
That's an excellent point. It seems likely they thought they could operate as a proper reviewer, but when the deadline came, they took the shortcut they knew they were not supposed to take.
It really does sound like an addiction when you put it this way.