Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Social Hippocampus Memory Learning

Liping Yi, Zhiming Zhao, Qinghua Hu

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.

Social Hippocampus Memory Learning

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.

Paper Structure

This paper contains 41 sections, 2 theorems, 25 equations, 7 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Social Learning and Social Machine Learning
  4. Heterogeneous Federated Learning
  5. Problem Definition
  6. The Proposed SoHip Algorithm
  7. Individual Short-Term Memory Abstraction
  8. Hippocampus-Inspired Memory Consolidation
  9. Individual--Collective Memory Fusion
  10. Collective Memory Aggregation
  11. Theoretical Analysis
  12. Experiments
  13. Experimental Setup
  14. Datasets and Data Partition.
  15. Models.
  16. ...and 26 more sections

Key Result

Theorem 5.1

Assume that each local objective $f_i$ is $L$-smooth and stochastic gradients have bounded variance. Under a suitable stepsize, the sequence generated by SoHip satisfies where $f(\theta)=\sum_{i=1}^N p_i f_i(\theta)$ and $\Delta_{\mathrm{het}}$ characterizes the effect of data and model heterogeneity across agents.

Figures (7)

  • Figure 1: Memory-centric social machine learning framework.
  • Figure 2: Overview of SoHip. SoHip operates sequentially by (1) abstracting individual short-term memory from local representations, (2) consolidating it into individual long-term memory via a hippocampus-inspired mechanism, (3) fusing it with collective long-term memory for enhanced prediction, and (4) aggregating updated individual long-term memories to form an updated collective memory.
  • Figure 3: Test accuracy curves under pathological label-skew settings ($C=10\%$) on CIFAR-100 (top) and Tiny-ImageNet (bottom) with varying numbers of agents, where SoHip consistently achieves faster convergence and higher final accuracy across all settings.
  • Figure 4: Impact of non-IID degree under pathological label-skew by varying classes per agent and practical label-skew by Dirichlet partition with concentration $\alpha$ ($\star$ is default in Table \ref{['tab:main_compare']}).
  • Figure 5: Impact of memory dimension $m$ on SoHip.
  • ...and 2 more figures

Theorems & Definitions (2)

  • Theorem 5.1: Convergence of SoHip
  • Theorem 5.2: Privacy Preservation