ArXiv, a widely used open repository for preprint research, is doing more to crack down on the careless use of large language models in scientific papers.
Although papers are posted to the site before they are peer-reviewed, arXiv (pronounced “archive”) has become one of the main ways that research circulates in fields like computer science and math, and the site itself has become a source of data on trends in scientific research.
ArXiv has already taken steps to combat a growing number of low-quality, AI-generated papers, for example by requiring first-time posters to get an endorsement from an established author. And after being hosted by Cornell for more than 20 years, the organization is becoming an independent nonprofit, which should allow it to raise more money to address issues like AI slop.
In its latest move, Thomas Dietterich — the chair of arXiv’s computer science section — posted Thursday that “if a submission contains incontrovertible evidence that the authors did not check the results of LLM generation, this means we can’t trust anything in the paper.”
That incontrovertible evidence could include things like “hallucinated references” and comments to or from the LLM, Dietterich said. If such evidence is found, a paper’s authors will face “a 1-year ban from arXiv followed by the requirement that subsequent arXiv submissions must first be accepted by a reputable peer-reviewed venue.”
