HE-OFT: Privacy-Preserving One-Shot Federated Fine-Tuning under Homomorphic Encryption

Key points

• Many organizations adapt large pretrained models to their own tasks by fine-tuning on private data.1
• Several of these parties often hold data for the same task and wish to fine-tune a model together without pooling that data.1
• Federated learning (FL) enables joint fine-tuning, but reconstruction attacks on shared intermediate values (the model or its gradients) remain a privacy risk.1
• A one-shot protocol that exchanges one encrypted contribution exposes no intermediate value.1

Sources (1)

• [1]HE-OFT: Privacy-Preserving One-Shot Federated Fine-Tuning under Homomorphic Encryption
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 6, 12:35 PM

  We present HE-OFT, the first cryptographically secure one-shot federated fine-tuning protocol in which no party receives the trained model.

  Many organizations adapt large pretrained models to their own tasks by fine-tuning on private data.

Extractive summary: sentences quoted from the sources.