Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
NovemberWhiskey
10 months ago
|
parent
|
context
|
favorite
| on:
CriticGPT: Finding GPT-4's mistakes with GPT-4
>
"outputting a sequence of tokens representing a claim that can be uncontroversially disproven," is still a useful concept.
Sure, but that would require semantic mechanisms rather than statistical ones.
riwsky
10 months ago
[–]
Statistics has a semantics all its own
Consider applying for YC's Summer 2025 batch! Applications are open till May 13
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
Sure, but that would require semantic mechanisms rather than statistical ones.