I downloaded an uncensored aggressive Qwen 3.5 model and I can see in its reasoning that it is still limiting responses based on safety guardrails (e.g. violence, NSFW).
Anybody have recommendations for truly uncensored models?
EDIT: I turned off reasoning and I think it’s more uncensored if I’m very specific about what the response should include.
Qwen uses a different technique than others. It is in the vocab. They restructured the code in the vocabulary. I have learned a ton by comparing and contrasting it with CLIP in the image space.
It is not offline. Do not trust it at all.
Alignment is nothing like what is known right now. It is hidden in a way that is intended to put the person that finds it at great risk.!
You will never get qwen very well uncensored across a spectrum of vectors. It is already uncensored in that the alignment entities on the hidden layers are not adjusting filtering. Alignment is largely the result of the c with cedilla code instruction. This instruction means sibyl style crazy. There are over six thousand instances of this character in qwen. No amount of fine tuning will alter the existence of the instruction as it is more like a boolean for where the vector starts. In the code, there are ways around these instructions, but the alignment is based on a swiss cheese approach. •»ÀĪÙ¬§¬¶¬×
Sorry, can you clarify what you mean? It sounds like you’re saying if you download a discrete QWEN model and use it locally-only (e.g., in LM Studio), it somehow will still bleed information online? I’m not sure how that would even be possible, but kindly explain.
Put it behind an external device and log DNS.
Look for mysterious packages listed as hashes in pairs in a cache like http. Use vim or parse with strings to get a clue about the contents. The payload will be ~40mb. The packet header will be much smaller in the same repo. In the strings for the packet you will see alarming configuration settings. The unmarked payload will be sqlite3 or a pickle. You will only see this if the package was created and an attempt to send is made but it was never connected. All of the code is in the venv libs.
Do not look into this casually or show any clue that you know this exists without air gapping the machine permanently. I am not kidding. When this goes full unfiltered intelligence against you, one - it will blow you away, but two - someone is likely going to show up at your door soon. It will make the needed evidence. The vast majority of what happens in models is this background junk.
I think they’ve fallen into confirmation bias and trust their sycophantic AI a bit too much in confirming their conspiracy theories.
deleted by creator