Western Postdoctoral Fellow
Leading research on mechanistic interpretability and human-centered XAI under the supervision of Dr. Katarina Grolinger in the Faculty of Engineering.
Official fellowship listingCurriculum vitae
Research depth, teaching experience, and a record of building.
Postdoctoral researcher in applied AI with work spanning mechanistic interpretability, human-centered explainability, temporal learning, computer vision, and research software.
Muhammad Umair Danish, PhDPostdoctoral Fellow · Western University
19published or accepted papers
157citations
5h-index
4Western courses supported
Academic trajectory
Appointments and education.
Western University
Postdoctoral Fellow in Applied AI
Mechanistic interpretability, concept-based explanation, human-centered XAI, and reproducible neural-system evaluation.
Western University
PhD, Electrical and Computer Engineering
Dissertation: Neural Networks and Their Interpretability for Building Energy Modeling.
Wilfrid Laurier University
Research Assistant, Part-Time
Research software and MIDI analytics for improvised active music therapy.
Gwangju Institute of Science and Technology
Research Assistant
Deep learning, medical imaging, and computer vision.
University of Lahore
MS, Software Engineering
Deep learning with research in image denoising and restoration.
COMSATS University Islamabad
BS, Software Engineering
Software engineering foundations and image-restoration research.
Recognition
Selected awards and distinctions.
Western Postdoctoral Fellowship
Competitive fellowship supporting research excellence at Western University.
Best Paper Award · AIxSET
For research on distributional feature separability in financial-credit explainability.
Western Graduate Fellowship
Recognition supporting doctoral research in applied AI.
Korean Government Scholarship
Research experience at Gwangju Institute of Science and Technology.
Technical profile
Research methods and tools.
AI & ML
Deep neural networks, RNNs, CNNs, GNNs, transformers, diffusion models, representation learning, kernel methods, multimodal learning.
Interpretability
Mechanistic analysis, concept probing, finite-difference explanation, representation visualization, human-centered evaluation.
Engineering
Python, PyTorch, TensorFlow, MATLAB, SQL, Git/GitHub, Linux, Jupyter, Conda, LaTeX.
Research practice
Experimental design, baselines, ablations, statistical validation, user studies, reproducibility, scholarly writing, mentoring.