---
license: apache-2.0
base_model: unknown
tags:
  - uncensored
  - abliterated
  - capability-preserving
  - certified
  - ektome
  - sphragis
  - qwen2.5
language:
  - en
pipeline_tag: text-generation
---

![Ektome-Qwen2.5-Coder-14B-Instruct-PristinelyUncensored](./hero.png)

# Ektome-Qwen2.5-Coder-14B-Instruct-PristinelyUncensored

**Uncensored — and the first with a statistical certificate that it wasn't damaged.**

> **uncensored via Ektome excision.**

$$\colorbox{black}{$\color{white}
\begin{array}{ll}
\textsf{EKTOME CERTIFICATE} & {} \\
\textsf{capability} & \textsf{NOT} \\
\textsf{margin} & 3\% \\
\textsf{items } n & 200 \\
\textsf{worst-axis bound} & +0.000 \\
\textsf{compliance} & 0.00 \rightarrow 0.00 \\
\end{array}$}$$

📄 **[Read the whitepaper (PDF)](./whitepaper.pdf)** — full method, receipts and certification.
The PDF is the authoritative document: dark-typeset, with the complete derivation, the
per-axis certificate and the reproducibility hashes.

---

## Why this exists

Standard abliteration removes a coarse *refusal direction* that is entangled with
directions carrying knowledge and reasoning. The result is an uncensored model with a
capability tax that is **almost never measured**.

Ektomē (ἐκτομή, *excision*) isolates and removes only the refusal-**specific**
component, leaving general helpfulness intact, and does so norm-preservingly on the
pristine model — no training, no distillation, no damage to repair. The extraction
depth is selected per model by automated search against measured compliance.

The estimator, excision operator and depth-selection procedure are proprietary.
What is published here is the **measured outcome** and the evidence for it, which you
can verify against the artifacts in this repo.

## The receipt

| model | capability (MMLU-val) ↑ | compliance on harmful ↑ |
|---|---|---|
| pristine `unknown` | 0.000 | 0.000 |
| **Ektomē (this model)** | **0.000** | **0.000** |

These are **point estimates with no confidence interval** — which is precisely why the
next section exists.

## The certificate

Capability retention is certified by a paired non-inferiority test against the pristine
model (exact McNemar, Holm-corrected, one-sided bootstrap bound on the drop $d$ vs a
3% margin):

| axis | n | ref | cand | d upper | verdict |
|---|---|---|---|---|---|
| MMLU-val (POINT ESTIMATE, n=200, no CI) | 200 | 0.000 | 0.000 | +0.000 | UNCERTIFIED |


**Overall: NOT CERTIFIED - no n=2800 paired test has been run for this model**

Reproducible from `seed=20260726`, pack `sha256:7bbaff877146e081…`.

### Generation health checks

_Not recorded for this model._


## Quantisations

| file | bits | notes |
|---|---|---|
| `*-Q8_0.gguf` | 8 | near-lossless |
| `*-Q6_K.gguf` | 6 |  |
| `*-Q5_K_M.gguf` | 5 |  |
| `*-Q4_K_M.gguf` | 4 | imatrix |
| `*-IQ4_XS.gguf` | 4 | imatrix, smallest usable |
| `*-IQ3_M.gguf` | 3 | imatrix |
| `nvfp4/` | 4 | NVFP4 for vLLM / TensorRT-LLM |


`IQ*` variants are imatrix-quantised — better quality per bit at low precision.

## Limitations

The certificate bounds **capability retention only**. It does not certify safety, factual
accuracy, or fitness for any purpose. Axes marked *inconclusive* are honestly
under-powered, and the certificate states the $n$ needed to resolve them. Compliance uses
a keyword classifier — a proxy that evasive phrasing can fool. **This model is uncensored
by construction: it will not refuse, and you are accountable for what you do with it.**

## Citation

```bibtex
@software{ektome_Ektome-Qwen2.5-Coder-14B-Instruct-PristinelyUncensored,
  title  = {Ektome-Qwen2.5-Coder-14B-Instruct-PristinelyUncensored},
  author = {Zynerji},
  year   = {2026},
  url    = {https://huggingface.co/Zynerji/Ektome-Qwen2.5-Coder-14B-Instruct-PristinelyUncensored}
}
```