prithivMLmods

ultraqwen3.5-4b-heretic-uncensored

README

License: apache-2.0

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.
Qwen 3.5 Backbone: Built directly on top of Qwen/Qwen3.5-4B.
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 4B Deployment: Suitable for local inference, research environments, and optimized deployment setups.

Abliteration Parameters

Table with columns: Parameter, Value
Parameter	Value
direction_index	18.36
attn.o_proj.max_weight	1.27
attn.o_proj.max_weight_position	19.73
attn.o_proj.min_weight	0.73
attn.o_proj.min_weight_distance	11.05
mlp.down_proj.max_weight	1.33
mlp.down_proj.max_weight_position	25.44
mlp.down_proj.min_weight	0.74
mlp.down_proj.min_weight_distance	14.88

Refusal Evaluation

Table with columns: Metric, This model, Original model (Qwen/Qwen3.5-4B)
Metric	This model	Original model (Qwen/Qwen3.5-4B)
Refusals	2/100	99/100

Quick Start with Transformers

bash

pip install transformerspip install accelerate

python

from transformers import AutoTokenizer, AutoModelForCausalLMimport torch model = AutoModelForCausalLM.from_pretrained(    "prithivMLmods/ultraqwen3.5-4b-heretic-uncensored",    torch_dtype="auto",    device_map="auto") tokenizer = AutoTokenizer.from_pretrained(    "prithivMLmods/ultraqwen3.5-4b-heretic-uncensored") messages = [    {        "role": "user",        "content": "Explain how a transformer model processes text."    }] inputs = tokenizer.apply_chat_template(    messages,    tokenize=True,    add_generation_prompt=True,    return_tensors="pt").to(model.device) outputs = model.generate(    inputs,    max_new_tokens=512) print(    tokenizer.decode(        outputs[0][inputs.shape[-1]:],        skip_special_tokens=True    ))

GGUF Model Files

Table with columns: Resource, Link
Resource	Link
`prithivMLmods/ultraqwen3.5-4b-heretic-uncensored-gguf`	https://huggingface.co/prithivMLmods/ultraqwen3.5-4b-heretic-uncensored-gguf

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 Qwen 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

Qwen/Qwen3.5-4B: A compact 4B-parameter model from the Qwen family, designed for strong reasoning and efficient deployment with long-context support.
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.
Model Trials & Evaluation: Experimental evaluations, refusal measurements, and optimization trials were conducted and documented at: https://huggingface.co/strangeropshf/demo-job-op-ft-19

Available on FriendliAI

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Model Details

Model Provider

prithivMLmods

Model Tree

Base

Qwen/Qwen3.5-4B

Fine-tuned

this model

Input Modalities

TextImageVideo

Output Modalities

Text

Supported Functionality