Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
JosephRedfern
16 days ago
|
parent
|
context
|
favorite
| on:
Show HN: 30k IKEA items in flat text
I've heard that LLMs can perform worse with these more efficient representations compared to e.g. JSON, because they've seen far fewer examples of them during training. Do you know how true that is?
TechSquidTV
16 days ago
|
next
[–]
Absolutely, but usually when working with a bespoke format for optimization, it's paired with an LLM specifically trained on that format.
tsazan
16 days ago
|
prev
[–]
You are right about cryptic formats. CommerceTXT is semantically structured. Models like GPT, Claude and Gemini understand it out-of-the-box via ICL.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: