Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
andyjohnson0
on Feb 19, 2023
|
parent
|
context
|
favorite
| on:
GPTZero Case Study – Exploring False Positives
Meta-ChatGPT's loss function optimises for ChatGPT generating training data that maximises the shittyness of Google's LLM.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
