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

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Guo, Tao, Wang, Junxiao, Huo, Fushuo, Cui, Laizhong, Guo, Song, Gui, Jie, Tao, Dacheng
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915456956235776
author Guo, Tao
Wang, Junxiao
Huo, Fushuo
Cui, Laizhong
Guo, Song
Gui, Jie
Tao, Dacheng
author_facet Guo, Tao
Wang, Junxiao
Huo, Fushuo
Cui, Laizhong
Guo, Song
Gui, Jie
Tao, Dacheng
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2508_16261
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On the Evolution of Federated Post-Training Large Language Models: A Model Accessibility View
Guo, Tao
Wang, Junxiao
Huo, Fushuo
Cui, Laizhong
Guo, Song
Gui, Jie
Tao, Dacheng
Machine Learning
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.
title On the Evolution of Federated Post-Training Large Language Models: A Model Accessibility View
topic Machine Learning
url https://arxiv.org/abs/2508.16261