Welcome to Francis Academic Press

The Frontiers of Society, Science and Technology, 2025, 7(8); doi: 10.25236/FSST.2025.070806.

Anniversaries as Periodic Re-seeding Events: A Generative-Agent Model of War Memory Diffusion and Decay

Author(s)

Shaobo Wu

Corresponding Author:
Shaobo Wu
Affiliation(s)

Business School, University of Shanghai for Science and Technology, Shanghai, China

Abstract

Research on collective memory has long distinguished communicative memory—interpersonal transmission through everyday interactions—from cultural memory, which is maintained by institutionalized rituals, symbolic practices, and annual commemorations. Yet existing empirical and theoretical work lacks a unified computational framework capable of simulating how these two channels interact to shape memory evolution over time. This paper proposes an AI-assisted agent-based modeling paradigm that models the dual structure of communicative and cultural memory and allows researchers to simulate the dynamic propagation, reinforcement, and decay of collective memory. We implement this framework in the context of war memory influenced by annual anniversaries. A small-world social network is populated with agents endowed with demographic and personality attributes. Communicative memory spreads as agents decide whether to discuss the anniversary with neighbors, and these decisions are generated by a large language model so that inter-agent variation approximates realistic human reasoning. Cultural memory appears as a daily exogenous activation process, with intensified activation on the anniversary date. The simulation shows that the model reproduces empirically observed features of real-world commemorative dynamics: sharp annual memory peaks, regular decay between anniversaries, and the amplification role of interpersonal clustering. Results demonstrate that integrating AI-driven decision-making with dual-channel memory structures produces realistic trajectories of collective remembrance. This approach offers a new methodological path for computational memory studies and contributes a flexible modeling paradigm that can be extended to disasters, pandemics, political events, and other domains where collective memory evolves through intertwined interpersonal and institutional processes.

Keywords

Agent-Based Simulation, Collective Memory, Commemorative Dynamics

Cite This Paper

Shaobo Wu. Anniversaries as Periodic Re-seeding Events: A Generative-Agent Model of War Memory Diffusion and Decay. The Frontiers of Society, Science and Technology (2025), Vol. 7, Issue 8: 34-40. https://doi.org/10.25236/FSST.2025.070806.

References

[1] Assmann J. Cultural memory and early civilization: Writing, remembrance, and political imagination[M]. Cambridge University Press, 2011. 

[2] Assmann J. Communicative and cultural memory[M]//Cultural memories: The geographical point of view. Dordrecht: Springer Netherlands, 2011: 15-27.

[3] Park J S, O'Brien J, Cai C J, et al. Generative agents: Interactive simulacra of human behavior[C]//Proceedings of the 36th annual acm symposium on user interface software and technology. 2023: 1-22.

[4] Watts D J. Strogatz-small world network Nature[J]. Nature, 1998, 393(June): 440-442.