Social Hippocampus Memory Learning

2026-03-26

AI Summary

This work addresses the challenges of privacy leakage and high communication overhead in heterogeneous federated learning, where sharing models or intermediate representations poses significant risks. To overcome these issues, the authors propose SoHip, a memory-centric social machine learning framework that introduces, for the first time, a hippocampus-inspired memory mechanism. In SoHip, each agent consolidates its local short-term memory into long-term memory and collaboratively learns through lightweight memory sharing and fusion, while keeping both data and models strictly local. This approach eliminates the need for model alignment and substantially reduces communication costs, offering provable theoretical privacy guarantees. Experimental results on two benchmark datasets demonstrate that SoHip significantly outperforms seven baseline methods, achieving a maximum accuracy improvement of 8.78%.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Humans and AI: Human-in-the-loop Machine Learning
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Security and Privacy: Security and privacy of machine learning and AI applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Social learning highlights that learning agents improve not in isolation, but through interaction and structured knowledge exchange with others. When introduced into machine learning, this principle gives rise to social machine learning (SML), where multiple agents collaboratively learn by sharing abstracted knowledge. Federated learning (FL) provides a natural collaboration substrate for this paradigm, yet existing heterogeneous FL approaches often rely on sharing model parameters or intermediate representations, which may expose sensitive information and incur additional overhead. In this work, we propose SoHip (Social Hippocampus Memory Learning), a memory-centric social machine learning framework that enables collaboration among heterogeneous agents via memory sharing rather than model sharing. SoHip abstracts each agent's individual short-term memory from local representations, consolidates it into individual long-term memory through a hippocampus-inspired mechanism, and fuses it with collectively aggregated long-term memory to enhance local prediction. Throughout the process, raw data and local models remain on-device, while only lightweight memory are exchanged. We provide theoretical analysis on convergence and privacy preservation properties. Experiments on two benchmark datasets with seven baselines demonstrate that SoHip consistently outperforms existing methods, achieving up to 8.78% accuracy improvements.

Problem

• federated learning
• heterogeneous agents
• privacy preservation
• model sharing
• social machine learning

Innovation

• Social Machine Learning
• Federated Learning
• Memory-Centric Learning
• Hippocampus-Inspired Mechanism
• Heterogeneous Agents