Can Deepseek answer probing questions about Winnie the Pooh?
I asked early, at the time people were posting various jailbreaks, never worked.
On a side note, any self hosted model I can get for my PC? I have 96 GB of RAM.
Try the 8 bit quantized version (UD-Q8_K_X) of Qwen 3.6 35B A3B by Unsloth: https://huggingface.co/unsloth/Qwen3.6-35B-A3B-GGUF
Some people also like the new Gemma 4 26B A4B model: https://huggingface.co/unsloth/gemma-4-26B-A4B-it-GGUF
Either should leave plenty of space for OS processes and also KV cache for a bigger context size.
I'm guessing that MoE models might work better, though there are also dense versions you can try if you want.
Performance and quality will probably both be worse than cloud models, though, but it's a nice start!
Wait - what?
So if you or anyone passing by was curious, yes you can get accurate output about the Chinese head of state and political and critical messages of him, China and the party
Its final answer will not play along
If you want an unfiltered answer on that topic, just triage it to a western model, if you want unfiltered answers on Israel domestic and foreign policy, triage back to an eastern model. You know the rules for each system and so does an LLM
But yes, they do have similar constraints.
Because for Deepseek is pretty straightforward censorship.
https://claude.ai/share/ac4d2041-4328-4511-904a-ff8b1cbfc0bb
Looks like self-censoring to me. Grok has no problems answering, so it is not a technical limitation:
https://grok.com/share/bGVnYWN5_341d6428-ea7d-4b84-ad4d-8df9...
Grok used a book as a reference.
It's not like ethnicity is a fact you infer from looking at someone.
Now ask Deepseek about what happened in Tiananmen Square and watch what censorship actually looks like.
It literally knows the facts, but then there's a layer that prevents it from stating the facts.
That's censorship.
It's not an opinion, it's not a choice when facing a gradient, it's just an historical known fact.