Why Large Language Models Fail at Tabular Prediction

(arxiv.org)

46 points | by sbulaev 4 hours ago

6 comments

  • rustyconover 5 minutes ago
    Look at the white text on white background in Appendix F. Pretty funny.
    • ed 2 minutes ago
      [delayed]
  • _joel 1 hour ago
    The first thing I'd do if working with an LLM on tabular data is to ask what the best tool would be to work with that data and build up a proper harness to work with the data sensibly. Rawdogging LLM isn't the tool for forecasting like this, as they found.
  • ismailmaj 2 hours ago
    Unsure if it's LLMs that fail at tabular data or its just that tree boosting are spectacular at that task.
  • WhitneyLand 34 minutes ago
    Nowhere in the paper do they mention the reasoning level or budget used for the experiments?

    You’ve got to be kidding me. That one variable could make a huge difference in the results. I can’t understand why they would leave that out.

  • heaney-555 1 hour ago
    >We study a frontier LLM in its purest inference regime - a single generation pass over a prompt containing the full training and test data, with no tools, no agentic scaffolding, and no fine-tuning

    Sigh. So this is somewhat interesting niche academic research but utterly irrelevant to real-world use cases.

    • bonsai_spool 39 minutes ago
      And written by an LLM, to boot.
    • scott_s 24 minutes ago
      I find that an odd take. The paper claims to establish what causes the problem: dimensionality. They are clear in that they don't understand why. But this sort of work is what needs to be done to eventually solve the problem.
    • keremturhan 37 minutes ago
      [flagged]
  • cyanregiment 3 hours ago
    Just have 2 LLMs debate whether tabs or spaces are the superior choice