davidemodolo/malicious_llm_finetuning

Proof of concept demonstrating backdoor injection into fine-tuned LLMs using LoRA. Shows how supposedly "open weights" models can be secretly compromised to harbor malicious behaviors triggered by specific patterns (email addresses).

22
/ 100
Experimental
No Package No Dependents
Maintenance 13 / 25
Adoption 0 / 25
Maturity 9 / 25
Community 0 / 25

How are scores calculated?

Stars

—

Forks

—

Language

Python

License

—

Category

membership-inference-attacks

Last pushed

Mar 18, 2026

Commits (30d)

0

GitHub
Membership Inference Attacks · 45 frameworks

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/davidemodolo/malicious_llm_finetuning"

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

Higher-rated alternatives

google/scaaml

SCAAML: Side Channel Attacks Assisted with Machine Learning

75

Koukyosyumei/AIJack

Security and Privacy Risk Simulator for Machine Learning (arXiv:2312.17667)

57

pralab/secml

A Python library for Secure and Explainable Machine Learning

54

AI-SDC/SACRO-ML

Collection of tools and resources for managing the statistical disclosure control of trained...

53

liuyugeng/ML-Doctor

Code for ML Doctor

45

Explore ML Frameworks

All categories Trending ML Framework directory Insights