Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
porridgeraisin
5 months ago
|
parent
|
context
|
favorite
| on:
A small number of samples can poison LLMs of any s...
It can. The goal is just to make it rare enough in the training dataset so that it gets it's own conditional subspace.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: