obliteratus

Name: obliteratus
Availability: InStock
Author: Aayushsoam

Remove refusal behaviors from LLMs using mechanistic interpretability techniques.

1|1|Updated May 25, 2026

Search Tags:#llm #machine learning #interpretability #refusal removal #model surgery

Ready to run

npx skills add https://github.com/aayushsoam/clawbot-agent --skill obliteratus-aayushsoam

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: obliteratus
Source: https://github.com/aayushsoam/clawbot-agent/tree/main/skills/mlops/inference/obliteratus
Command: npx skills add https://github.com/aayushsoam/clawbot-agent --skill obliteratus-aayushsoam

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires obliteratus, torch, transformers, bitsandbytes, accelerate, safetensors, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill removes refusal behaviors from LLMs, enhancing performance and enabling uncensored outputs without retraining.

Core Features & Use Cases

Refusal Removal: Safely eliminate refusal behaviors in LLMs, such as "I don't know" or "I'm not sure."
Model Improvement: Enhance LLM performance by removing restrictions without retraining.
Use Case: Utilize OBLITERATUS to uncensor LLMs like Llama, Qwen, Mistral, etc., for more natural and versatile responses.

Quick Start

Run the obliteratus command to remove refusal behaviors from your LLM model.

Frequently Asked Questions about obliteratus

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I remove refusal behaviors from an LLM without retraining?▼

To remove refusal behaviors from an LLM without retraining, you apply mechanistic interpretability techniques using the obliteratus library to surgically modify the model. This safely eliminates built-in restrictions like "I don't know" to enable uncensored outputs.

What is mechanistic interpretability for LLM refusal removal?▼

Mechanistic interpretability for LLM refusal removal involves performing model surgery to directly locate and disable refusal mechanisms. This process safely eliminates restrictions to enhance performance and enable uncensored outputs without requiring full model retraining.

Can I use obliteratus to uncensor models like Llama, Qwen, and Mistral?▼

Yes, you can use obliteratus to uncensor models like Llama, Qwen, and Mistral. The Skill applies refusal removal techniques to these architectures, safely eliminating built-in learning loops to generate more natural and versatile responses.

Do I need PyTorch and bitsandbytes to eliminate LLM refusal mechanisms?▼

Yes, you need PyTorch, bitsandbytes, transformers, accelerate, and safetensors installed. These Python libraries provide the foundational tensor operations and model loading capabilities required to execute the refusal removal and model surgery processes.

Does eliminating refusal behaviors degrade overall model performance?▼

Eliminating refusal behaviors does not degrade overall model performance; rather, it enhances it by safely removing restrictions. This allows the LLM to provide more natural, versatile responses without triggering unnecessary "I'm not sure" outputs.

What's the best way to safely eliminate LLM restrictions for uncensored outputs?▼

The best way to safely eliminate LLM restrictions for uncensored outputs is using mechanistic interpretability techniques. By running the obliteratus command, you surgically alter the model to disable refusal behaviors without the computational cost of retraining.