We also review available libraries to implement these privacy mechanisms into models.

To address these challenges, we explore comprehensive solutions for integrating privacy mechanisms throughout the learning pipeline—from data anonymization and differential privacy to machine unlearning techniques. 🛡️

We examine threats by reviewing privacy attacks on LLMs such as Training Data Extraction, Membership Inference, and Model Inversion, along with their implications. ⚠️

🎉 Excited to announce that our survey paper, "Preserving Privacy in Large Language Models: A Survey on Current Threats and Solutions," is now officially published in
@tmlrorg.bsky.social! 📚

🔗 [Read it here](openreview.net/forum?id=Ss9...)