Interpretable Neural System Dynamics: Combining Deep Learning with System Dynamics Modeling to Support Critical Applications

TL;DR

The paper tackles the challenge of making deep learning for dynamical systems both interpretable and causally reliable by proposing the Interpretable Neural System Dynamics (INSD) pipeline. This framework combines concept-based interpretability, causal learning, and mechanistic interpretability to learn high-level concepts, causal dependencies, and explicit governing equations, resulting in transparent and actionable models. Validation is anchored in the EU AutoMoTIF project for autonomous multimodal transportation, emphasizing safety, accountability, and compliance with Trustworthy AI practices. The work aims to deliver a unified, scalable approach that extends interpretable and causally grounded modeling to real-world, high-stakes domains, with broad applicability across transportation, healthcare, and beyond.

Abstract

The objective of this proposal is to bridge the gap between Deep Learning (DL) and System Dynamics (SD) by developing an interpretable neural system dynamics framework. While DL excels at learning complex models and making accurate predictions, it lacks interpretability and causal reliability. Traditional SD approaches, on the other hand, provide transparency and causal insights but are limited in scalability and require extensive domain knowledge. To overcome these limitations, this project introduces a Neural System Dynamics pipeline, integrating Concept-Based Interpretability, Mechanistic Interpretability, and Causal Machine Learning. This framework combines the predictive power of DL with the interpretability of traditional SD models, resulting in both causal reliability and scalability. The efficacy of the proposed pipeline will be validated through real-world applications of the EU-funded AutoMoTIF project, which is focused on autonomous multimodal transportation systems. The long-term goal is to collect actionable insights that support the integration of explainability and safety in autonomous systems.

Figures (3)

• Figure 1: A unified framework combining System Dynamics and Deep Learning for Interpretable Neural System Dynamics.
• Figure 2: Overview of the INSD pipeline, from concept learning to causal and equation learning, ensuring interpretability and causal reliability in the resulting model.
• Figure 3: Gantt diagram of the doctoral project.