Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
federico-peconi
34 days ago
|
parent
|
context
|
favorite
| on:
A small number of samples can poison LLMs of any s...
Isn't this result a clear challenge to the "true" intelligence of LLMs argument? Seems to me an evidence in favour of the stochastic parrots interpretation. Am I missing something?
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: