p-e-w / heretic
Training-free language model modification: Remove censorship, suppress slop, reduce sycophancy, and more
Why this job: works in the areas we cover, but does not say which job it does
What it does
Heretic is a unique tool that lets you do what only corporations with access to massive computing power usually do: *Modify,* rather than merely run, language models. It does not perform expensive training; rather, it uses a collection of mathematical techniques to steer behavior based on contrasting example prompts and matching scorers.
Get started
pip install -U heretic-llm
heretic Qwen/Qwen3-4B-Instruct-2507 - Maintenance
- Active
- Last commit 2 days ago
- Stars
- 40,366
- +778 this week
- Licence
- AGPL-3.0
- Share-alike, even hosted
- Downloads for
- None listed
- No release files
- Latest version
- v2.1.0
- Released 4 days ago
- Built with
- Python
- 100% of the code
Is it looked after?
- Contributors
- 13
- Last 90 days
- Issues
- 86% closed
- 34 still open
- Pull requests
- 143 merged
- 29 waiting
- Releases
- 7
- 7 in the last year
- No contributing guide
- No security policy
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| lordx64 / phantom-kv | 274 | 52 days ago | MIT | No downloads |
| Orion-zhen / abliteration | 203 | 2 days ago | GPL-3.0 | No downloads |
More: languages, topics and project record
Languages
- Python 100%
Topics
- abliteration
- llm
- transformer
- censorship-resistance
- censorship
- anti-censorship
- anticensorship
Project record
- JobOther AI tools
- StatusFeatured
- Commits239
- Forks4,627
- Watching205
- Stars in 28 days+5,031
- CreatedSep 21, 2025
- First seen by usNov 7, 2026
- CheckedNov 12, 2026