Research Focus

My current research interests lie in enhancing structured object representation through two complementary approaches: neuro-symbolic reasoning and integrating structured physics principles into machine learning frameworks. With particular applications towards ab initio physical sciences and generalizing physics-based intelligence. 
 

Education

• PhD Student, Machine Learning
  University of Basel (2021 – Present)

• Ex-PhD Student, Physics
  University of Luxembourg (2019 – 2021)

• Master of Science, Bioinformatics
  Boston University (2016 – 2018)

• Bachelor of Science, Biology
  Michigan State University (2011 – 2016)
   

Selected Publications

1. Diamond, J., & Lill, M. A. (2024). Neural SHAKE: Geometric Constraints in Graph Generative Models. International Conference on Artificial Neural Networks, 43-57.

2. Diamond, J., & Lill, M. (2023). Geometric Constraints in Probabilistic Manifolds: A Bridge from Molecular Dynamics to Structured Diffusion Processes. 2023 ICML Workshop. arXiv:2307.04493.

3. Diamond, J., & Lill, M. (2023). Improving Protein-peptide Interface Predictions in the Low Data Regime. 2023 ICLR Workshop on Drug Discovery. arXiv:2306.00557.