HackerTrans
TopNewTrendsCommentsPastAskShowJobs

mattcollins

no profile record

Submissions

United Arab Emirates to quit oil cartel OPEC

bbc.co.uk
15 points·by mattcollins·hace 2 meses·1 comments

A Timeline to China Blocking Meta's $2B Manus Acquisition (Built Using Manus)

metamanus-rsbcnkpx.manus.space
3 points·by mattcollins·hace 3 meses·0 comments

US Justice Department releasing more than three million pages from Epstein files

bbc.co.uk
41 points·by mattcollins·hace 5 meses·13 comments

Which Nested Data Format Do LLMs Understand Best? JSON vs. YAML vs. XML vs. MD

improvingagents.com
2 points·by mattcollins·hace 9 meses·1 comments

comments

mattcollins
·hace 4 meses·discuss
On the other hand, AI coding tools make it relatively easy to set and apply policies that can help with this sort of thing.

I like to have something like the following in AGENTS.md:

## Guiding Principles - Optimise for long-term maintainability - KISS - YAGNI
mattcollins
·hace 9 meses·discuss
Results from some further tests here: https://www.improvingagents.com/blog/toon-benchmarks
mattcollins
·hace 9 meses·discuss
FWIW, I ran a test comparing LLM accuracy with TOON versus JSON, CSV and a variety of other formats when using them to represent tabular data: https://www.improvingagents.com/blog/is-toon-good-for-table-...

I've only looked at one model (gpt-4.1-nano) so far. I'm hoping to run similar tests on some other models but it gets challenging to discern statistically significant differences with better models as their accuracy tends to be a lot better across the board.
mattcollins
·hace 9 meses·discuss
This is a follow-up to previous work looking at which format of TABULAR data LLMs understand best: https://www.improvingagents.com/blog/best-input-data-format-...

(There was some good discussion on Hacker News around that here: https://news.ycombinator.com/item?id=45458455)

We often want to feed NON-TABULAR data to LLMs, though, such as typical API responses or config files.

This new work looks out how the format of such nested / hierarchical data affects how well LLMs can answer questions about it; specifically how several models get on with JSON, YAML, XML and Markdown.
mattcollins
·hace 9 meses·discuss
Here you go: https://www.improvingagents.com/blog/best-input-data-format-...
mattcollins
·hace 9 meses·discuss
Author here.

This has made me chuckle several times - thanks!
mattcollins
·hace 9 meses·discuss
I did a small test with just a couple of formats and something like 100 records, saw that the accuracy was higher than I wanted, then increased the number of records until the accuracy was down to 50%-ish (e.g. 100 -> 200 -> 500 -> 1000, though I forget the precise numbers.)
mattcollins
·hace 9 meses·discuss
I'm the person who ran the test.

To hopefully clarify a bit...

I intentionally chose input data large enough that the LLM would be scoring in the region of 50% accuracy in order to maximise the discriminative power of the test.
mattcollins
·hace 9 meses·discuss
I'm the person who ran the test.

To explain the 60% a bit more...

With small amounts of input data, the accuracy is near 100%. As you increase the size of the input data, the accuracy gradually decreases.

For this test, I intentionally chose an input data set large enough that the LLM would score in the region of 50% accuracy (with variation between formats) in order to maximise the discriminative power of the test.
mattcollins
·hace 9 meses·discuss
I'm the person who ran the test.

The context I used in the test was pretty large. You'll see much better (near 100%) accuracy if you're using smaller amounts of context.

[I chose the context size so that the LLM would be scoring in the ballpark of 50% accuracy (with variation between formats) to maximise the discriminative power of the test.]