Zhang-Yihao/Adversarial-Representation-Engineering
Official implementation repository for the paper Towards General Conceptual Model Editing via Adversarial Representation Engineering.
This project helps AI researchers and developers modify the behavior of large language models (LLMs). It takes an LLM and specific instructions for desired behavioral changes (e.g., reducing harmful responses or hallucinations), then outputs a modified LLM that adheres to these new guidelines. This tool is for those who are building or deploying LLMs and need to fine-tune their safety and accuracy.
No commits in the last 6 months.
Use this if you are a researcher or developer working with large language models and need to specifically control their output behavior, such as minimizing harmful content or factual errors.
Not ideal if you are an end-user of an AI application or are looking for a no-code solution to general LLM fine-tuning.
Stars
19
Forks
2
Language
Python
License
MIT
Category
Last pushed
Dec 06, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/Zhang-Yihao/Adversarial-Representation-Engineering"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
MadryLab/context-cite
Attribute (or cite) statements generated by LLMs back to in-context information.
microsoft/augmented-interpretable-models
Interpretable and efficient predictors using pre-trained language models. Scikit-learn compatible.
Trustworthy-ML-Lab/CB-LLMs
[ICLR 25] A novel framework for building intrinsically interpretable LLMs with...
poloclub/LLM-Attributor
LLM Attributor: Attribute LLM's Generated Text to Training Data
THUDM/LongCite
LongCite: Enabling LLMs to Generate Fine-grained Citations in Long-context QA