Front Propagation

Front propagation describes the spread of information or influence across a network, whether it's a neural network or a graph representing relationships. Research focuses on developing efficient algorithms, such as front-propagation for explainable AI and F-adjoint for clarifying backpropagation in neural networks, to model and understand this spread. These models find applications in diverse fields, including explainable AI, network analysis, and semi-supervised learning, by providing insights into information flow and decision-making processes within complex systems. The unification of different front propagation models, including wave-based and eikonal-based approaches, is also a key area of current investigation.