HenryCai11/LLM-Self-Control

The official repo of paper "Self-Control of LLM Behaviors by Compressing Suffix Gradient into Prefix Controller"

31
/ 100
Emerging

This project helps people who work with Large Language Models (LLMs) fine-tune their behavior. You provide a rule or desired behavior as a text string, and the system automatically adjusts the LLM's responses to follow that rule. This means you can get more consistent and controlled outputs without manually labeling data, making it useful for AI developers or researchers.

No commits in the last 6 months.

Use this if you need to reliably guide an LLM to produce specific types of responses, such as making it sound more emotional, ensuring it avoids harmful content, or improving its reasoning without changing its core programming.

Not ideal if you need a solution for a small, simple LLM, or if you prefer to use traditional fine-tuning methods with extensive human-annotated datasets.

LLM-customization AI-safety content-moderation persona-control response-biasing
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 6 / 25
Maturity 16 / 25
Community 9 / 25

How are scores calculated?

Stars

18

Forks

2

Language

Jupyter Notebook

License

MIT

Last pushed

Aug 13, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/HenryCai11/LLM-Self-Control"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.