A Decoupled Alignment Kernel for Peptide Membrane Permeability Predictions
arXiv, 2025
Ali Amirahmadi
Industrial Postdoctoral Researcher at AstraZeneca
Generative Models · Language Models · Machine Learning for Science
I am an industrial postdoctoral researcher at AstraZeneca working at the intersection of machine learning, generative AI, and molecular science.
My research focuses on developing computational methods for complex scientific data, with particular interest in generative models, language models, molecular structure elucidation, and machine learning for scientific discovery.
I am especially interested in representation learning, multimodal modeling, and generative approaches for molecular and structured data.
Research Interests
- Generative Models Generative methods for scientific and molecular data, including representation learning and probabilistic modeling.
- Language Models Language and foundation models for scientific reasoning, structured information, and domain-specific applications.
- Machine Learning for Science Developing data-driven methods for molecular, biomedical, and other scientific applications.
Selected Publications
View all publications →Group-Sparse Manifold-Aware Integrated Gradients for Multimodal Transformers on EHR Trajectories
Machine Learning for Health (ML4H), 2025