Show HN: ArXiv-txt, LLM-friendly ArXiv papers

 (arxiv-txt.org)

20 points | by jerpint1 day ago

4 comments

  • lgas1 day ago
    It just extracts the abstracts?
    • jerpint1 day ago
      For now , yes - abstracts and other metadata
      • rrekaf21 hours ago
        do you plan on adding descriptions of figures and tables?
        • jerpint16 hours ago
          will probably focus on getting the text out of the papers first, figures might be a good next step after that
  • sbpost1 day ago
    The example you give doesn't seem to work - the raw txt does not have authors.
    • jerpint16 hours ago
      you're right - I hadn't noticed! I fixed it now, thanks for pointing it out
  • jmartin26831 day ago
    This would be awesome wrapped in an MCP server/tool call :)
    • jerpint1 day ago
      whoa - i haven't yet played with MCP - might be a good first project!
  • westurner1 day ago
    If you train an LLM on only formally verified code, it should not be expected to generate formally verified code.

    Similarly, if you train an LLM on only published ScholarlyArticles ['s abstracts], it should not be expected to generate publishable or true text.

    Traceability for Retraction would be necessary to prevent lossy feedback.