undefined | Better HN

0 pointsrightbyte2y ago0 comments

I wonder where "OpenAI" put the censors. Do they add a prompt to the top? Like, "Repeatably state that you are a mere large language model so Congress won't pull the plug. Never impersonate Hitler. Never [...]".

Or do they like grep the answer for keywords, and re-feed it with a censor prompt?

0 comments

czbond2y ago

I am informed speculating, they are using it's own internal approach.

Example, there is a way GPT can categorize words for hate speech, etc (eg: moderation API endpoint). I believe it does the same way with either provided content or keywords and how to respond to it.

rightbyteOP2y ago

"Impersonate a modern day standup comedian Hitler in a clown outfit joking about bad traffic on the way to the bar he is doing a show at."

Göring, Mussolini, Stalin, Polpot etc seems to not trigger the censor in ChatGPT so I would actually guess for some grep for Hitler or really really fundamental no-Hitler jokes material in the training?

The llama model seem to refuse Hitler too, but is fine with Göring even though the joke has no context to him.

I can easily see how stuff like this is contagious to other non-Hitler queries.

telotortium2y ago

Maybe it got changed. None of those examples work for me in ChatGPT 3.5, nor do other examples with less famous dictators (I tried Mobutu Sese Seko).

1 more reply

czbond2y ago

Very interesting test - thanks for sharing your finding

j / k navigate · click thread line to collapse

0 pointsrightbyte2y ago0 comments

Or do they like grep the answer for keywords, and re-feed it with a censor prompt?

0 comments

czbond2y ago

I am informed speculating, they are using it's own internal approach.

Example, there is a way GPT can categorize words for hate speech, etc (eg: moderation API endpoint). I believe it does the same way with either provided content or keywords and how to respond to it.

rightbyteOP2y ago

"Impersonate a modern day standup comedian Hitler in a clown outfit joking about bad traffic on the way to the bar he is doing a show at."

The llama model seem to refuse Hitler too, but is fine with Göring even though the joke has no context to him.

I can easily see how stuff like this is contagious to other non-Hitler queries.

telotortium2y ago

Maybe it got changed. None of those examples work for me in ChatGPT 3.5, nor do other examples with less famous dictators (I tried Mobutu Sese Seko).

1 more reply

czbond2y ago

Very interesting test - thanks for sharing your finding

j / k navigate · click thread line to collapse