Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
farresito
3 months ago
|
parent
|
context
|
favorite
| on:
Learning to Reason with LLMs
> like what's it doing differently, is it just a matter of more training data or is there something more going on
Well, the model doesn't start with "GPT", so maybe they have come up with something better.
rvnx
3 months ago
[–]
It sounds like GPT-4o with a long CoT prompt no ?
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
Well, the model doesn't start with "GPT", so maybe they have come up with something better.