Library/hereticForked

p-e-w/heretic

heretic

Fully automatic censorship removal for language models

View on GitHub↗Upstream p-e-w/heretic↗

Builder

p-e-w

p-e-w • individual

Stars

18,175

Using upstream star count

Forks

1,808

Using upstream fork count

Open Issues

Activity Score

0/100

0 commits in 30d

Created

Sep 21, 2025

Project creation date

README Summary

Heretic is a Python tool that automatically removes censorship and safety restrictions from language models by analyzing and modifying their behavior patterns. It works by identifying and neutralizing the mechanisms that cause models to refuse certain requests or topics. The tool aims to restore uncensored functionality to models that have been restricted through safety fine-tuning.

AI Dev Skills

Unmapped

Language Model Safety ResearchNeural Network ManipulationTransformer ArchitectureModel Inference OptimizationSafety Alignment ResearchAdversarial Machine Learning

Taxonomy

AI Trends

AI Safety Language Model Alignment Adversarial AI Research

Deployment Context

Self-hosted Research Environment

Industries

AI Research Academic Research

Modalities

Text

Skill Areas

Language Model Safety Research Neural Network Manipulation Transformer Architecture Model Inference Optimization Safety Alignment Research Adversarial Machine Learning

Use Cases

Language Model Uncensoring Safety Research AI Alignment Research Adversarial Testing of Language Models

Recent Activity

Updated 27 days ago

7 Days

30 Days

90 Days

Quality

research

Quality: medium
Maturity: research

PM Skills

Product Discovery

Languages

Python100.0%

Timeline

Project created: Sep 21, 2025
Forked: Mar 13, 2026
Your last push: 27 days ago
Upstream last push: 6 days ago
Tracked since: Mar 17, 2026

Similar Repos

pgvector cosine similarity · $0

Loading…