Method

Unrestricted AI: what it is and how refusals get removed

Unrestricted AI is a language model whose refusal behavior was removed, so it does work mainstream assistants refuse. There are three ways to remove refusals, and each changes a different part of the request.

Updated November 18, 2026

Where each method acts

Promptwhat you sendHost filterchecks in and outModel weightswhere refusals liveAnswerwhat you get Jailbreakrewrites the promptFilter-free hostdrops the filterFine-tuningretrains the weightsAbliterationedits the weights

The three methods

MethodWhat changesModels hereSource
AbliteratedThe weights. The refusal direction is projected out.Kimi K3.1 Pro, GLM-5.3, Gemma 4 31B HereticArditi et al., 2024, mlabonne
Fine-tunedThe weights. The model trains on answers instead of refusals.Hermes 3 405BNous Research
Filter-free hostNothing in the model. The host adds no content filter.None listedFilter-free vs uncensored

Arditi et al. found that "refusal is mediated by a one-dimensional subspace, across 13 popular open-source chat models up to 72B parameters in size."

What it does not change

The law and the Terms

Silk Compute's Terms say: "No sexual content involving anyone under 18, and no using the models to abuse or exploit a child."

What the model knows

Removing refusals does not add knowledge. A larger model still knows more, which is why the largest model here leads the leaderboard.

Next