Why Large Language Models Fail at Tabular Prediction

Look at the white text on white background in Appendix F. Pretty funny.

_joel an hour ago | parent | prev | next [-]

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 | parent | prev | next [-]

Unsure if it's LLMs that fail at tabular data or its just that tree boosting are spectacular at that task.

	▲	dist-epoch 35 minutes ago \| parent [-]
		Non-LLM transformers beat tree boosting - TabPFN, TabFM https://research.google/blog/introducing-tabfm-a-zero-shot-f...

▲

WhitneyLand 29 minutes ago | parent | prev | next [-]

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 an hour ago | parent | prev | next [-]

>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.

	▲	scott_s 19 minutes ago \| parent \| next [-]
		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.
	▲	bonsai_spool 34 minutes ago \| parent \| prev [-]
		And written by an LLM, to boot.

▲

cyanregiment 3 hours ago | parent | prev [-]

Just have 2 LLMs debate whether tabs or spaces are the superior choice