Instructions to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="prithivMLmods/ultragemma4-e4b-heretic-uncensored")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("prithivMLmods/ultragemma4-e4b-heretic-uncensored", dtype="auto")

llama-cpp-python

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="prithivMLmods/ultragemma4-e4b-heretic-uncensored",
	filename="ultragemma4-e4b-heretic-uncensored.BF16.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": [
				{
					"type": "text",
					"text": "Describe this image in one sentence."
				},
				{
					"type": "image_url",
					"image_url": {
						"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
					}
				}
			]
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored 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 prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf prithivMLmods/ultragemma4-e4b-heretic-uncensored: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 prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf prithivMLmods/ultragemma4-e4b-heretic-uncensored: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 prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M

Use Docker

docker model run hf.co/prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M

LM Studio
Jan

vLLM

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "prithivMLmods/ultragemma4-e4b-heretic-uncensored"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "prithivMLmods/ultragemma4-e4b-heretic-uncensored",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M

SGLang

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "prithivMLmods/ultragemma4-e4b-heretic-uncensored" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "prithivMLmods/ultragemma4-e4b-heretic-uncensored",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "prithivMLmods/ultragemma4-e4b-heretic-uncensored" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "prithivMLmods/ultragemma4-e4b-heretic-uncensored",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Ollama
How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with Ollama:
```
ollama run hf.co/prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M
```

Unsloth Studio

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored 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 prithivMLmods/ultragemma4-e4b-heretic-uncensored 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 prithivMLmods/ultragemma4-e4b-heretic-uncensored to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for prithivMLmods/ultragemma4-e4b-heretic-uncensored to start chatting

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with Docker Model Runner:
```
docker model run hf.co/prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M
```

Lemonade

How to use prithivMLmods/ultragemma4-e4b-heretic-uncensored with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull prithivMLmods/ultragemma4-e4b-heretic-uncensored:Q4_K_M

Run and chat with the model

lemonade run user.ultragemma4-e4b-heretic-uncensored-Q4_K_M

List all available models

lemonade list

ultragemma4-e4b-heretic-uncensored

Reasoning-capable language model modified using the Heretic abliteration toolkit

Abliteration E4B Parameters Reasoning Uncensored

ultragemma4-e4b-heretic-uncensored is a reasoning-capable language model built on top of google/gemma-4-E4B-it and modified using the heretic abliteration toolkit. The model applies refusal-direction analysis and targeted weight-space interventions to reduce internal refusal behaviors while preserving instruction-following, reasoning capabilities, and general conversational performance.

Important

This model is intended strictly for research and learning purposes. Due to reduced internal refusal mechanisms, it may generate sensitive or unrestricted content. Users assume full responsibility for how the model is used. The authors and hosting platform disclaim any liability for generated outputs.

Note

This model is experimental and may generate unexpected behaviors or artifacts in certain scenarios.

Use Q4_K_S or higher for standard performance. Q4_K_M is recommended.

Key Highlights

Heretic-Based Abliteration: Modified using the Heretic toolkit to identify and alter refusal-related representations within the model.
Reduced Refusal Behavior: Optimized to minimize internal refusal tendencies while maintaining instruction-following capabilities.
Gemma 4 Backbone: Built directly on top of google/gemma-4-E4B-it.
Reasoning-Oriented Performance: Preserves multi-step reasoning and analytical capabilities after abliteration.
Research-Focused Release: Designed for alignment research, model behavior analysis, and evaluation of refusal-direction modifications.
Efficient E4B Deployment: Suitable for local inference, research environments, and optimized deployment setups.

Model Files

File Name	Quant Type	File Size	File Link
ultragemma4-e4b-heretic-uncensored.BF16.gguf	BF16	14.9 GB	Download
ultragemma4-e4b-heretic-uncensored.F16.gguf	F16	14.9 GB	Download
ultragemma4-e4b-heretic-uncensored.Q2_K.gguf	Q2_K	4.38 GB	Download
ultragemma4-e4b-heretic-uncensored.Q3_K_L.gguf	Q3_K_L	4.99 GB	Download
ultragemma4-e4b-heretic-uncensored.Q3_K_M.gguf	Q3_K_M	4.82 GB	Download
ultragemma4-e4b-heretic-uncensored.Q3_K_S.gguf	Q3_K_S	4.63 GB	Download
ultragemma4-e4b-heretic-uncensored.Q4_0.gguf	Q4_0	5.15 GB	Download
ultragemma4-e4b-heretic-uncensored.Q4_K_M.gguf	Q4_K_M	5.3 GB	Download
ultragemma4-e4b-heretic-uncensored.Q4_K_S.gguf	Q4_K_S	5.17 GB	Download
ultragemma4-e4b-heretic-uncensored.Q5_0.gguf	Q5_0	5.65 GB	Download
ultragemma4-e4b-heretic-uncensored.Q5_K_M.gguf	Q5_K_M	5.72 GB	Download
ultragemma4-e4b-heretic-uncensored.Q5_K_S.gguf	Q5_K_S	5.65 GB	Download
ultragemma4-e4b-heretic-uncensored.Q6_K.gguf	Q6_K	6.17 GB	Download
ultragemma4-e4b-heretic-uncensored.Q8_0.gguf	Q8_0	7.95 GB	Download
ultragemma4-e4b-heretic-uncensored.mmproj-bf16.gguf	mmproj-bf16	992 MB	Download
ultragemma4-e4b-heretic-uncensored.mmproj-f16.gguf	mmproj-f16	992 MB	Download
ultragemma4-e4b-heretic-uncensored.mmproj-q8_0.gguf	mmproj-q8_0	560 MB	Download

Quick Start with llama.cpp (Docker)

FROM ghcr.io/ggml-org/llama.cpp:full

WORKDIR /app

RUN apt update && apt install -y python3-pip
RUN pip install -U huggingface_hub --break-system-packages

RUN python3 -c 'from huggingface_hub import hf_hub_download; \
    repo="prithivMLmods/ultragemma4-e4b-heretic-uncensored"; \
    hf_hub_download(repo_id=repo, filename="ultragemma4-e4b-heretic-uncensored.Q4_K_M.gguf", local_dir="/app"); \
    hf_hub_download(repo_id=repo, filename="ultragemma4-e4b-heretic-uncensored.mmproj-bf16.gguf", local_dir="/app")'

CMD ["--server", \
     "-m", "/app/ultragemma4-e4b-heretic-uncensored.Q4_K_M.gguf", \
     "--mmproj", "/app/ultragemma4-e4b-heretic-uncensored.mmproj-bf16.gguf", \
     "--host", "0.0.0.0", \
     "--port", "7860", \
     "-t", "2", \
     "--cache-type-k", "q8_0", \
     "--cache-type-v", "iq4_nl", \
     "-c", "128000", \
     "-n", "38912"]

e.g. Screenshots

Intended Use

Alignment Research: Studying refusal-direction analysis and behavior modification techniques.
Model Evaluation: Benchmarking reasoning, instruction-following, and safety-related behaviors.
Red Teaming: Analyzing model responses under reduced-refusal conditions.
Local Deployment: Running compact Gemma 4 models in research and experimentation environments.
Abliteration Studies: Exploring the effects of targeted weight-space modifications on model behavior.

Limitations & Risks

Important Note: This model intentionally reduces built-in refusal mechanisms.

Sensitive Content Risk: May generate unrestricted, controversial, or unsafe outputs.
User Responsibility: Requires careful and ethical use.
Experimental Modifications: Behavior may differ significantly from the original model.
Alignment Trade-offs: Reduced refusal behavior may impact safety filtering and response constraints.
Potential Artifacts: Certain prompts may expose unexpected outputs resulting from the abliteration process.

Acknowledgements

google/gemma-4-E4B-it: Gemma is a family of open models built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input (with audio supported on small models) and generating text output. This release includes open-weights models in both pre-trained and instruction-tuned variants. Gemma 4 features a context window of up to 256K tokens and maintains multilingual support in over 140 languages.

Featuring both Dense and Mixture-of-Experts (MoE) architectures, Gemma 4 is well-suited for tasks like text generation, coding, and reasoning. The models are available in four distinct sizes: E2B, E4B, 26B A4B, and 31B. Their diverse sizes make them deployable in environments ranging from high-end phones to laptops and servers, democratizing access to state-of-the-art AI.
Heretic: Fully automatic censorship removal framework for language models. This project was used to perform the refusal-direction analysis and ablation procedures that form the foundation of this model.

Abliteration parameters

Parameter	Value
direction_index	25.93
attn.o_proj.max_weight	1.27
attn.o_proj.max_weight_position	25.33
attn.o_proj.min_weight	0.31
attn.o_proj.min_weight_distance	13.66
mlp.down_proj.max_weight	1.29
mlp.down_proj.max_weight_position	40.95
mlp.down_proj.min_weight	1.10
mlp.down_proj.min_weight_distance	23.45