Lachlan Simpson

Lachlan Simpson

Higher Degree by Research Candidate

School of Electrical and Mechanical Engineering

Faculty of Sciences, Engineering and Technology

My research leverages tools for differential geometry and Lie theory to develop robust and user-friendly explainable AI models. I also work in developing physics-informed neural networks (PINNs) for problems in wave physics. In particular, I have applied PINNs to quantum graphs to engineer desirable properties of lattices. 

  • Education
    Date Institution name Country Title
    University of Adelaide Australia Bachelor of Computer and Mathematical Sciences
    University of Adelaide Australia Honours Bachelor of Computer and Mathematical Sciences (Pure Mathematics)

  • Book Chapters
    Year Citation
    2025 Costanza, E., & Simpson, L. (2025). Riemannian Integrated Gradients: A Geometric View of Explainable AI. In F. Nielsen, & F. Barbaresco (Eds.), Geometric Science of Information 7th International Conference, GSI 2025, Saint-Malo, France, October 29–31, 2025, Proceedings, Part I. Springer Nature.

  • Conference Papers
    Year Citation
    2024 Simpson, L., Costanza, F., Millar, K., Cheng, A., Lim, C. C., & Chew, H. G. (2024). Algebraic Adversarial Attacks on Integrated Gradients. In Proceedings - International Conference on Machine Learning and Cybernetics (pp. 26-31). Hybrid, Miyazaki: IEEE.
    DOI
    2024 Simpson, L., Costanza, F., Millar, K., Cheng, A., Lim, C. C., & Chew, H. G. (2024). Tangentially Aligned Integrated Gradients for User-Friendly Explanations. In CEUR Workshop Proceedings Vol. 3910 (pp. 1-12). Dublin, Ireland: CEUR-WS.
    2023 Simpson, L., Millar, K., Cheng, A., Chew, H. G., & Lim, C. C. (2023). A Testbed for Automating and Analysing Mobile Devices and Their Applications. In Proceedings - International Conference on Machine Learning and Cybernetics (pp. 201-208). Online: IEEE.
    DOI Scopus1
    2022 Millar, K., Simpson, L., Cheng, A., Chew, H. G., & Lim, C. (2022). Detecting Botnet Victims Through Graph-Based Machine Learning. In 2021 International Conference on Machine Learning and Cybernetics (ICMLC) Vol. 2021-December (pp. 6 pages). online: IEEE.
    DOI Scopus1

  • Preprint
    Year Citation
    2025 Simpson, L., Millar, K., Cheng, A., Lim, C. C., & Chew, H. G. (2025). Graph-based Integrated Gradients for Explaining Graph Neural Networks.
    DOI
    2025 Simpson, L., Costanza, F., Millar, K., Cheng, A., Lim, C. -C., & Chew, H. G. (2025). Algebraic Adversarial Attacks on Explainability Models.
    2024 Simpson, L., Millar, K., Cheng, A., Lim, C. -C., & Chew, H. G. (2024). Probabilistic Lipschitzness and the Stable Rank for Comparing
    Explanation Models.

ANA Partnership Grant Scheme: Physics-Informed Neural Networks for Quantum Graph-Based Medical Imaging Devices

Connect With Me
External Profiles