# AbliteratedOutTheBox Strata Announcement — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/2e1e8c59b19d8f54

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-10-24 13:38 ingested 2026-10-25 02:39

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xe8x9c/abliteratedoutthebox_strata_announcement/)

## retained text



```
I’ve been working on an inference-time abliteration engine as a fork of llama.cpp, and I’ve now ported the hook to Strata.

It acts as an easy way to use abliterated models without huge amounts of vram or finetuning.

For the Strata Branch

[https://github.com/heterodoxin/AbliteratedOutTheBox/tree/strata](https://github.com/heterodoxin/AbliteratedOutTheBox/tree/strata)

For the llama.cpp branch

[https://github.com/heterodoxin/AbliteratedOutTheBox](https://github.com/heterodoxin/AbliteratedOutTheBox)

A star or upvote would help this project a lot.

The Strata port is experimental, full-model behavior and response quality still need validation. If anyone running Strata wants to try it, I’d appreciate feedback
```



document fetched 2026-10-25 00:36 · canonical https://reddit.com/r/LocalLLaMA/comments/1xe8x9c/abliteratedoutthebox_strata_announcement
