On the Evolution of Federated Post-Training Large Language Models: A Model Accessibility View

TL;DR

This paper surveys federated tuning of large language models, focusing on different levels of model access and efficiency, and discusses the emerging black-box inference paradigm in federated learning.

Contribution

It provides a comprehensive taxonomy of federated LLM tuning methods based on model access and efficiency, highlighting the black-box inference approach and future challenges.

Findings

Classifies FedLLM approaches into white-box, gray-box, and black-box categories.

Highlights the shift towards inference-only black-box methods in federated LLM tuning.

Discusses open challenges and future research directions in federated LLMs.

Abstract

Federated Learning (FL) enables training models across decentralized data silos while preserving client data privacy. Recent research has explored efficient methods for post-training large language models (LLMs) within FL to address computational and communication challenges. While existing approaches often rely on access to LLMs' internal information, which is frequently restricted in real-world scenarios, an inference-only paradigm (black-box FedLLM) has emerged to address these limitations. This paper presents a comprehensive survey on federated tuning for LLMs. We propose a taxonomy categorizing existing studies along two axes: model access-based and parameter efficiency-based optimization. We classify FedLLM approaches into white-box, gray-box, and black-box techniques, highlighting representative methods within each category. We review emerging research treating LLMs as black-box inference APIs and discuss promising directions and open challenges for future research.