arxivcs.AIcs.DL2026-07-23
Traceable Scholarship: Page Anchors and Ariadne's Thread for Humanistic Inquiry in the Age of Generative AI
Generative AI lets large language models produce scholarly-looking text within seconds, yet fluency does not equal valid explanation. The deepest risk is not factual error alone but the appearance that an explanation is already established without clear sources, page numbers, edi…