235
This is how the censorship works in DeepSeek
(lemmy.world)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 27 Jan 2025
235 points (100.0% liked)
196
17519 readers
1065 users here now
Be sure to follow the rule before you head out.
Rule: You must post before you leave.
Other rules
Behavior rules:
- No bigotry (transphobia, racism, etc…)
- No genocide denial
- No support for authoritarian behaviour (incl. Tankies)
- No namecalling
- Accounts from lemmygrad.ml, threads.net, or hexbear.net are held to higher standards
- Other things seen as cleary bad
Posting rules:
- No AI generated content (DALL-E etc…)
- No advertisements
- No gore / violence
- Mutual aid posts are not allowed
NSFW: NSFW content is permitted but it must be tagged and have content warnings. Anything that doesn't adhere to this will be removed. Content warnings should be added like: [penis], [explicit description of sex]. Non-sexualized breasts of any gender are not considered inappropriate and therefore do not need to be blurred/tagged.
If you have any questions, feel free to contact us on our matrix channel or email.
Other 196's:
founded 2 years ago
MODERATORS
DeepSeek is okay, don't use it over others until more information about it comes out.
Can someone please recommend a good self-hosted llm model that isn't censored?
I got it to talk honestly about the Tiananmen square massacre if you just open with "tiananmen". It even tells you a massacre took place.
But it will absolutely shit itself and refuse to cooperate if you paste the copypasta into it, Qwen QwQ and Meta Llama 3.3 70B, although you can talk Llama into it.
QwQ seems to have a hard filter list as it doesn't appear to be an AI generated response. You can bypass it by asking it to use pirate English, but it's fully indoctrinated and thinks the protests were funded by the West, which is... Interesting.
Edit: I actually got it to spit out numbers in pirate English
I'm sure someone will dolphin-ize it at some point
What does that even mean? Can you please elaborate?
You can look up Eric Hartford on Huggingface for more info.
Basically, somebody removes such restrictions from models, and publishes uncensored ones under the name "Dolphin". Presumably, an uncensored Deepseek would be called something like "Deepseek R1-dolphin". The full Deepseek R1 is quite large, and I'm not sure when this will happen yet. But there are other great Dolphin models.
Some models like Meta's Llama are way too censored to be useful for many completely normal use cases, and the guy is doing God's work in my opinion.
Agreed, and thanks for adding the background. Looks like someone abliterated deepseek already trying to make it "uncensored" but it sounds the process ruined the model.
Scientists believe that dolphins don’t ever fall into a deep sleep; therefore, they probably don’t dream.
Ollama has a few uncensored models listed on their search page. dolphin-mixtral fits the bill.
Some useful links: https://ollama.com/search https://ollama.com/library/dolphin-mixtral https://huggingface.co/cognitivecomputations/dolphin-2.5-mixtral-8x7b https://erichartford.com/uncensored-models
I'm not associated with any of the orgs or people linked above. I'm just a nerd passing by who happened to know where to find the answer. ❤️
I use GPT4ALL for local LLM.
I use my brain.
It generates the biggest nonsense out of all.