PluralLLM: Pluralistic Alignment in LLMs via Federated Learning
Abstract
Ensuring Large Language Models (LLMs) align with diverse human preferences while preserving privacy and fairness remains a challenge. Existing methods, such as Reinforcement Learning from Human Feedback (RLHF), rely on centralized data collection, making them computationally expensive and privacy-invasive. We introduce PluralLLM1 a federated learning-based approach that enables multiple user groups to collaboratively train a transformer-based preference predictor without sharing sensitive data, which can also serve as a reward model for aligning LLMs. Our method leverages Federated Averaging (FedAvg) to aggregate preference updates efficiently, achieving 46% faster convergence, a 4% improvement in alignment scores, and nearly the same group fairness measure as in centralized training. Evaluated on a Q/A preference alignment task, PluralLLM demonstrates that federated preference learning offers a scalable and privacy-preserving alternative for aligning LLMs with diverse human values.
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.