A tool that removes censorship from open-weight LLMs

Are there LLMs which don't always approve whatever idea the user has and tell him it's absolutely brilliant?

That doesn't mean there couldn't be a "concept neuron" that is doing the vast majority of heavy lifting for content refusal, though.

This is not what an ablation study is. An ablation study removes and/or swaps out ("ablates") different components of an architecture (be it a layer or set of layers, all activation functions, backbone, some fixed processing step, or any other component or set of components) and/or in some cases other aspects of training (perhaps a unique / different loss function, perhaps a specialized pre-training or fine-tuning step, etc) in order to attempt to better understand which component(s) of some novel approach is/are actually responsible for any observed improvements. It is a very broad research term of art.

That being said, the "Ablation Strategies" [1] the repo uses, and doing a Ctrl+F for "ablation" in the README does not fill me with confidence that the kind of ablation being done here is really achieving what the author claims. All the "ablation" techniques seem "Novel" in his table [2], i.e. they are unpublished / maybe not publicly or carefully tested, and could easily not work at all.

From later tables, I am not convinced I would want to use these ablations, as they ablate rather huge portions of the models, and so probably do result in massively broken models (as some commenters have noted in this thread elsewhere). EDIT: Also, in other cases [1], they ablate (zero out) architecture components in a way that just seems incredibly braindead if you have even a basic understanding of the linear algebra and dependencies between components of a transformer LLM. There is nothing sound clearly about this, in contrast to e.g. abliteration [3].

[1] hhtps://github.com/elder-plinius/OBLITERATUS?tab=readme-ov-file#ablation-strategies

[2] https://github.com/elder-plinius/OBLITERATUS?tab=readme-ov-f...

EDIT: As another user mentions, "ablation" has a specific additional narrower meaning in some refusal analyses or when looking at making guardrails / changing response vectors and such. It is just a specific kind of ablation, and really should actually be called "abliteration", not "ablation" [3].

[3] https://huggingface.co/blog/mlabonne/abliteration, https://arxiv.org/abs/2512.13655.

Also, as I said in a top level comment, what this project wants to achieve has been done for a while and it's called Heretic: https://github.com/p-e-w/heretic

(Not vibecode by a twitter influgrifter)

I use Berkley Sterling from 2024 because I can trick it. No abliteration needed.

    Strategy            What it does  Use case
    .......................................................
    layer_removal       Zero out      entire transformer layers
    head_pruning        Zero out      individual attention heads
    ffn_ablation        Zero out      feed-forward blocks
    embedding_ablation  Zero out      embedding dimension ranges

It's interesting that people are writing tools that go inside the weights and do things. We're getting past the black box era of LLMs.

That may or may not be a good thing.

Its not a frontier model but it will give you a feel for what its like.

Probably not, because if it is completely uncensored, it would probably violate the law (in different ways) in every possible jurisdiction.. (Also, one common method of censorship is exclusion of particular types of content from the training set, so to be completely free of that kind of censorship, there would have to be no content intentionally excluded from the training set.)

In general, paid services are censored not only to attempt to meet the laws in all jurisdictions of concern to the provider, but also to try to be safe with regard to the (shiifting) demands of payment processors, and to try to maintain the PR image of the provider.