[chmod775]

> there's an uncensored model that you can run locally with llama.cpp

Correction: There's tens of thousands of them. They're easy to create, which is why everyone publishes their own.

Just put "uncensored", "abliterated", or "heretic" into search on huggingface/ollama/etc and pick any them. Fair warning: most aren't very good, essentially lobotomized, and totally broken if you enable thinking.

[Scaled]

In my own tests, the abliterated models perform equivalently to the same version in an apples-to-apples comparison (if you compare same quantization). Thinking is working also. The main difference is I don't get annoying prompt refusals (otherwise common due to my work on 18+ related projects). However, it's local quantized models so they're not anywhere near frontier quality.

[desterothx]

interesting, I haven't played with any of them yet, but i thought the point of orthogonalizing the weights towards the restriction vector was that there is no loss in capability while removing guardrails. Does it affect other parts of the RL alignment too?

[embedding-shape]

> the point of orthogonalizing the weights towards the restriction vector was that there is no loss in capability while removing guardrails

That is surely the point, most of the "uncensored" weights released for free on HuggingFace aren't being very successful at this. There is a stark difference in output quality between the official weights and all these "uncensored" variants that appears days afterwards.

[fy20]

The claims by the creators are it doesn't in a major way. I have a uncensored Gemma 4 I run on my Mac. Just for testing out, I haven't found any need for it... yet.

[larodi]

The name for it is ablation - precise removal of parts of the model. Not abliteration as it is not obliteration.

Even as I write this the ‘abliterated’ word is denoted a typo. Does it not at your end?

[gliptic]

The name for it _is_ abliterate. It's a portmanteau of ablate and obliterate.

[larodi]

No tis not. The terminology has medical origins, and it is very much this

https://en.wikipedia.org/wiki/Ablation

"Ablation (Latin: ablatio – removal) is the removal or destruction of something from an object by vaporization, chipping, erosive processes, or by other means."

which is what one does to the model during... well, ablation.

[gliptic]

Except we're talking about a _different_ word, a word that was _coined_ as a portmanteau of the word you're referring to and another word.

[larodi]

Rillii ease eat? R U sure

https://en.wikipedia.org/wiki/Ablation_(artificial_intellige...

Have you read some (any) actual papers from the word this all and LLMs are?

Well here some:

https://arxiv.org/abs/1901.08644

Tis ablation ablation and ablation everywhere save for some kid’s fav. model names and their X pals.

[gliptic]

The very wikipedia page you're linking says:

> The term abliteration has been coined for the process of using ablation to uncensor large language models by modifying internal functions to completely eliminate refusal behaviors while preserving the remaining functions of the model. The word is a portmanteau that combines the words ablation and obliteration.

So are you dropping this now?

[eggsyntax]

Papers using the term in their title or abstract: https://arxiv.org/search/cs?query=abliteration&searchtype=al...

Note that they're all talking about the same technique, namely applying ablation specifically to the refusal vector as described in https://arxiv.org/abs/2406.11717.

I really recommend letting this one go.

[zargon]

The name is abliterate. It's a specific method of ablation.

[SonOfLilit]

https://github.com/NousResearch/llm-abliteration

[larodi]

well see my other comment. with all due respect to Teknium, perhaps he misspelled it.

[gliptic]

https://en.wiktionary.org/wiki/abliterate

[larodi]

first version...

curprev 22:29, 29 May 2025 ZeevoX talk contribs 2,320 bytes +2,320 Create page for "abliterate" Tag: added link

not very credible.

[SonOfLilit]

I wnt and read your other comments.

Except for the initial comment, they were aggressive and sometimes disrespectful claims that "the word is spelled ablation, here are some papers that use it" in response to people saying "yes, we know what ablation means, but GP is intentionally using a separate word 'abliteration' that is the accepted word for the kind of ablation he's talking about, see links to respectable sources using or defining it". I downvoted them because I feel the discussion would be more valuable and feel nicer to read without them, and you could just read any of the offered links before responding and save the trouble.

[larodi]

Is surely super agressive to outright downvote given evidence says otherwise but that’s fine I guess. Let’s drop it, apparently we’re not getting to mutually satisfying conclusion.

[larodi]

Tell these guys they’re wrong

https://arxiv.org/abs/1901.08644

[3abiton]

I miss the days of 4changpt.