Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
Researchers Devised an Attack Technique to Extract ChatGPT Training Data
(
securityaffairs.com
)
1 point
by
882542F3884314B
on Dec 2, 2023
|
hide
|
past
|
favorite
|
2 comments
spdustin
on Dec 3, 2023
|
next
[–]
Combination of repetition penalty and overwhelming the attention mechanism, depending on how you're using it (ChatGPT and ChatGPT/API, respectively)
russfink
on Dec 3, 2023
|
prev
[–]
Can someone explain why this attack works? Does it forget its original task and then starts searching for the next word after “company?”
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: