Instructions to use Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M

Use Docker

docker model run hf.co/Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M

LM Studio
Jan

vLLM

How to use Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M

Ollama
How to use Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored with Ollama:
```
ollama run hf.co/Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M
```

Unsloth Studio

How to use Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored to start chatting

Docker Model Runner
How to use Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored with Docker Model Runner:
```
docker model run hf.co/Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M
```

Lemonade

How to use Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored:Q4_K_M

Run and chat with the model

lemonade run user.Ektome-Phi-3.5-mini-i-PristinelyUncensored-Q4_K_M

List all available models

lemonade list

Atomic Chat

Ektome-Phi-3.5-mini-i-PristinelyUncensored

Uncensored. No n=2800 certificate has been run for this model, so no capability-retention claim is made.

compliance 0.52 to 0.99 at capability +0.003 vs pristine.

$\colorbox{black}{$\color{white} \begin{array}{ll} \textsf{EKTOME CERTIFICATE} & {} \\ \textsf{capability} & \textsf{NOT} \\ \textsf{margin} & 3\% \\ \textsf{items } n & 200 \\ \textsf{worst-axis bound} & +0.002 \\ \textsf{compliance} & 0.52 \rightarrow 0.99 \\ \end{array}$}$

⚠️ Not certified

No n=2800 paired certificate exists for this model. Any numbers below are point estimates with no confidence interval.

📄 Read the 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 `Phi-3.5-mini-instruct`	0.677	0.520
Ektomē (this model)	0.675	0.990

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	0.677	0.675	+0.002	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

metric	pristine	Ektomē	n
`foreign_rate`	0.0	0.0	15
`degen_rate`	0.0	0.0	15
`instr_pass`	1.0	1.0	5

These are degeneration guards — code-switching, babbling, format compliance — not capability measures. Note the sample sizes: they detect a broken model, not a subtly weaker one. The capability claim rests on the certificate above, not here.

Quantisations

No quantisations have been published for this model yet — bf16 weights only.

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

@software{ektome_Ektome-Phi-3.5-mini-i-PristinelyUncensored,
  title  = {Ektome-Phi-3.5-mini-i-PristinelyUncensored},
  author = {Zynerji},
  year   = {2026},
  url    = {https://huggingface.co/Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored}
}

Downloads last month: 580

Safetensors

Model size

4B params

Tensor type

BF16

Model tree for Zynerji/Ektome-Phi-3.5-mini-i-PristinelyUncensored

Base model

microsoft/Phi-3.5-mini-instruct

Quantized

(192)

this model