WebsiteHunt is broughtBrought to you by Redact Everything
Heretic: censorship removal for LMs

Heretic: censorship removal for LMs

Automatic censorship removal for language models.

Heretic removes censorship from transformer-based language models without post-training. It uses directional ablation with a TPE-based optimizer to decensor models while preserving performance.

Comments

Loading comments...

You might also like

HiAPI

HiAPI

One API for leading AI image, video, and text generation models. ✨ Premium
ProjectManagementTools

ProjectManagementTools

Optimize your Projects with ProjectManagementTools 🌱 Sponsored
ChatGPT

ChatGPT

Optimizing language models for dialogue
StableLM

StableLM

Stability AI's language models
OpenCoder

OpenCoder

The Open Cookbook For Top-Tier Code Large Language Models
VideoPoet

VideoPoet

A large language model for zero-shot video generation
Replit Code V1.5 3B

Replit Code V1.5 3B

A new code generation language model
Introducing DBRX

Introducing DBRX

Revolutionizing Language Models with DBRX
Language Learning with Netflix

Language Learning with Netflix

Use movies as study material for language learning
LibreLingo

LibreLingo

Learn a language for free
CA

Carton

Run any ML model from any programming language.
LL

Local LLM

Run large language models locally with practical guidance.
LLaMA

LLaMA

A foundational, 65-billion-parameter large language model
Language Atlas

Language Atlas

Learn a new language for free
Molmo by Ai2

Molmo by Ai2

Multimodal Open Language Model built by Ai2
Ollama

Ollama

The easiest way to run large language models locally
Smarter Quiz

Smarter Quiz

Are you smarter than a language model?
MH

Marko: HTML-based UI language

Declarative HTML-based language for fast, reactive apps.
Webctl

Webctl

CLI browser automation for AI agents
git-absorb

git-absorb

Automatic git commit fixups for developers
PaLM 2

PaLM 2

Google's next generation large language model
Lamini

Lamini

LLM engine for rapidly customizing models
by @Micadep