Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Too much censorship in Korean (BLOCKLIST, or OTHERS) #615

Open
hee-dongdong opened this issue Oct 31, 2024 · 0 comments
Open

Too much censorship in Korean (BLOCKLIST, or OTHERS) #615

hee-dongdong opened this issue Oct 31, 2024 · 0 comments
Labels
component:other Questions unrelated to SDK status:triaged Issue/PR triaged to the corresponding sub-team type:help Support-related issues

Comments

@hee-dongdong
Copy link

hee-dongdong commented Oct 31, 2024

Description of the bug:

"자인이", "유나", "혜인" those kinds of words are excessively censored. Actually it is just normal korean name. Moreover, When it comes to the sentence like "이 디자인이 저에게는 호감이네요." (which means 'I prefer this design') the LLM model just censor that entire sentence because of "자인이". even though that is not a human name in the context of the sentence.
This is a very critical problem and you need to fix it.
It could not be fixed by just tweaking safety settings or prompt

Actual vs expected behavior:

"이 디자인이 저에게는 호감이네요"
"사람들의 진짜 존경하는 것은, 그의 지혜인 것이었다."

Above normal sentences are censored, just because the words "자인이" and "혜인" are included.

Any other information you'd like to share?

No response

@gmKeshari gmKeshari added status:triaged Issue/PR triaged to the corresponding sub-team type:help Support-related issues component:other Questions unrelated to SDK labels Nov 4, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
component:other Questions unrelated to SDK status:triaged Issue/PR triaged to the corresponding sub-team type:help Support-related issues
Projects
None yet
Development

No branches or pull requests

2 participants