Exclusive: the science papers that patents cite the most

TL;DR - A Nature news feature (05 Aug 2026) reporting an exclusive analysis of which scientific papers are cited most often in patent filings, and the reasons behind their outsized industrial pull. Only the headline and abstract-level blurb were provided, so specifics below are what can be inferred rather than reported findings.

• Framed as a Nature journalism piece (doi:10.1038/d41586-026-02386-8), not a peer-reviewed study — it presents a ranked list of the most patent-cited research papers.
• Uses patent-to-paper citation linkage as a proxy for the science-to-technology transfer pathway, a common bibliometric signal of commercial/applied impact.
• Relevant to AI readers because foundational method papers (e.g., deep learning and computational tooling) typically dominate such rankings, though the provided content does not name any specific papers.
• Content is thin: no methodology, dataset, time window, or actual paper rankings were included in the supplied text — the full article would be needed to verify claims.