arXiv has implemented stricter moderation that rejects submissions like position papers and literature reviews for low quality or incorrect citations sparking debate on preprint openness
Rejections include genuine human citation errors from overworked researchers.
@tdietterich @JessicaHullman We all agree on that, we are just afraid that many good actors are going to be caught in the middle for honest editing mistakes.
If the bar is set in a way to only flag clear slop, nobady will have anything to argue.
Thomas G. Dietterich@tdietterich@JessicaHullman The majority of the papers where we find hallucinations are also single-author, "independent researchers". Many others are from paper mills or sock puppet accounts trying to boost citation counts. You don't want this on arXiv; it is a threat to knowledge.
@tdietterich I agree, I don't want it on arxiv! But I do see a difference between integrity filters like reject for fake citations & epistemic filters like reject if evidence of AI. The first rejects papers lacking basic legibility. The second asserts a stronger claim about what rigor means.
Thomas G. Dietterich@tdietterich@JessicaHullman The majority of the papers where we find hallucinations are also single-author, "independent researchers". Many others are from paper mills or sock puppet accounts trying to boost citation counts. You don't want this on arXiv; it is a threat to knowledge.