Michael Spratling

Dr. Michael Spratling is a Postdoctoral Researcher in the Department of Behavioural and Cognitive Sciences at the University of Luxembourg. His work bridges computational neuroscience and artificial intelligence, with specialization in neural networks, computer vision, and biologically-inspired learning systems. Spratling develops robust AI models through adversarial training techniques and neural architecture innovations.

His core research investigates predictive coding mechanisms in machine learning, adversarial defenses for computer vision systems, and biologically-plausible neural network architectures. Recent projects focus on improving model resilience against data perturbations using adaptive smoothing techniques.

Spratling maintains active GitHub repositories featuring benchmark tools for evaluating deep learning robustness. His work has applications in medical imaging and autonomous systems requiring reliable visual recognition. He received the Vebleo Scientist Award in 2020 for contributions to molecular biology-inspired AI research.

He collaborates internationally on projects involving few-shot anomaly detection and multimodal neural networks. Current grants support research in length extrapolation for large language models and margin-based loss functions.