Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
shikon7
56 days ago
|
parent
|
context
|
favorite
| on:
Heretic: Automatic censorship removal for language...
It seems to me that thinking models are harder to decensor, as they are trained to think whether to accept your request.
int_19h
55 days ago
[–]
It goes both ways. E.g. unmodified thinking Qwen is actually easier to jailbreak to talk about things like Tiananmen by convincing it that it is unethical to refuse to do so.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: