Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
CjHuber
3 months ago
|
parent
|
context
|
favorite
| on:
A small number of samples can poison LLMs of any s...
Imagine someone contaminated their training data into believing they are rational thinking machines
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
