Alexander Wang

I am a student in the Faculty of Mathematics the University of Cambridge, with an interest in statistics, machine learning and computational biology. I am member of Sidney Sussex College.

My research as an AI Scientist at Boehringer Ingelheim focuses on developing single-cell foundation models for understanding complex biological systems, in particular for cancer biology and single-cell transcriptomics.

I was previously a ML Researcher in the Artificial Intelligence Group under Professor Mateja Jamnik and Mateo Espinosa Zarlenga, working on interpretable machine learning. My research focused on concept based models, developing architectures for spatial concept localisation in images. We also developed interactive user tools that allowed interventions for concept based models.

My research interests include:

• Computational biology
• Single-cell genomics and transcriptomics
• Explainable AI
• Representation learning
• Foundation models

Within Computational Biology, I am interested in single-cell transcriptomics and the development of foundation models that learn representations of cellular state from gene expression data. A central question is how these representations can be used to support downstream biological tasks, such as characterising cell populations, modelling cellular interactions, predicting cellular responses to gene perturbations in order to understand disease progression.

I am also interested in the development of virtual cell models that integrate information across multiple biological processes, including gene regulation, signalling pathways, cellular communication, and tissue organisation. The long-term goal of this work is to build computational models capable of predicting cellular behaviour under novel conditions.

My research in Explainable AI focuses on concept-based models, which seek to represent predictions in terms of human-understandable concepts rather than opaque latent features, with the aim of improving the reliability of deep neural networks. I am particularly interested in how concept representations can improve model transparency by providing insight into the underlying reasoning processes and also enabling effective human intervention during inference.

- Sep 2023 – July 2026: B.A. in Mathematics, University of Cambridge

- Jul 2026 – Sep 2026: Junior AI Scientist, Boehringer Ingelheim

- Jun 2025 – Oct 2025: Machine Learning Researcher, Artificial Intelligence Group, University of Cambridge

- Jul 2024 – Sep 2024: Quantitative Researcher, Huatai-PineBridge Fund Management

- Jul 2023 – Aug 2023: Program in Mathematics for Young Scientists (PROMYS), University of Oxford