aiqbench analysis
FedCoT addresses the critical bottleneck of communication overhead in federated learning by enabling local generation of reasoning chains while aggregating only lightweight modules on the server. This approach allows large language models to maintain interpretable, auditable decision-making processes in privacy-sensitive domains like healthcare without relying on centralized data distillation. The method effectively balances accuracy with resource efficiency, offering a viable path forward for deploying advanced AI reasoning in distributed environments.
FedCoT: Communication-Efficient Federated Reasoning Enhancement for Large Language Models
arXiv:2508.10020v2 Announce Type: replace Abstract: Enhancing LLM reasoning in federated settings is nontrivial due to stringent computational, communication, and privacy constraints, especially in healthcare, where clinically consequential decisions require not only accuracy but also interpretable, auditable rationales to meet safety, accountability, and regulatory requirements. Conventional federated fine-tuning largely imitates final answers rather than cultivating step-by-step reasoning, often relying on privacy-sensitive centralized distillation and still incurring substantial communication overhead. We address this gap with \textbf{\ours{}}, a federated reasoning framework that combines lightweight chain-of-thought resampling with a compact discriminator for selection, and client-aware LoRA stacking with weighted classifier aggregation to accommodate heterogeneity while reducing aggregation noise and communication; clients generate candidate chains and supervision locally, and only lightweight modules are aggregated on the server. Experiments on medical reasoning benchmarks show consistent gains under tight resource budgets while keeping data local and respecting privacy, offering an interpretable and resource-efficient solution. Our code is made publicly available at https://github.com/DIaacKr/FedCoT
Abstract:Enhancing LLM reasoning in federated settings is nontrivial due to stringent computational, communication, and privacy constraints, especially in healthcare, where clinically consequential decisions require not only accuracy but also interpretable, auditable rationales to meet safety, accountability, and regulatory requirements. Conventional federated fine-tuning largely imitates final answers rather than cultivating step-by-step reasoning, often relying on privacy-sensitive centralized distillation and still incurring substantial communication overhead. We address this gap with \textbf{\ours{}}, a federated reasoning framework that combines lightweight chain-of-thought resampling with a compact discriminator for selection, and client-aware LoRA stacking with weighted classifier aggregation to accommodate heterogeneity while reducing aggregation noise and communication; clients generate candidate chains and supervision locally, and only lightweight modules are aggregated on the server. Experiments on medical reasoning benchmarks show consistent gains under tight resource budgets while keeping data local and respecting privacy, offering an interpretable and resource-efficient solution. Our code is made publicly available at this https URL
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
arXiv:2508.10020 [cs.CL]
(or arXiv:2508.10020v2 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2508.10020 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 248 words
Related coverage
Why I Left Google DeepMind
LessWrong · Jul 15, 2026
News
Quantum Computing Report · Jul 25, 2026
ParaBlock: Communication-Computation Parallel Block Coordinate Federated Learning for Large Language Models
arXiv · Jun 3, 2026
Spur: Modern GPU Job Scheduling for HPC and AI Workloads
AMD · Jul 22, 2026