English

A differentiable model of supply-chain shocks

Physics and Society 2025-11-10 v1 Machine Learning Multiagent Systems

Abstract

Modelling how shocks propagate in supply chains is an increasingly important challenge in economics. Its relevance has been highlighted in recent years by events such as Covid-19 and the Russian invasion of Ukraine. Agent-based models (ABMs) are a promising approach for this problem. However, calibrating them is hard. We show empirically that it is possible to achieve speed ups of over 3 orders of magnitude when calibrating ABMs of supply networks by running them on GPUs and using automatic differentiation, compared to non-differentiable baselines. This opens the door to scaling ABMs to model the whole global supply network.

Cite

@article{arxiv.2511.05231,
  title  = {A differentiable model of supply-chain shocks},
  author = {Saad Hamid and José Moran and Luca Mungo and Arnau Quera-Bofarull and Sebastian Towers},
  journal= {arXiv preprint arXiv:2511.05231},
  year   = {2025}
}

Comments

Accepted to 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Workshop: Differentiable Systems and Scientific Machine Learning (EurIPS)

R2 v1 2026-07-01T07:26:07.056Z