Perceptron Explanator (2020). Researchers and software engineers have been developing deeper artificial neural networks. This deepening augments the capabilities of these systems, but, as a consequence, also raises the number of calculations or decision iterations. These increments, in the number of iterations, give way to an increasing statistical complexity in the decision-making process, turning these systems into black boxes.
Perceptron Explanator is an interactive and algorithmic visualization that reveals the accumulated decision iterations of the most elementary system of an artificial neural network: a perceptron. It explains the decision-making progress, revealing the causality present in the process, while freeing the observer from the statistical complexity intrinsic to the algorithm. This visualization aims to help develop an intuition about the inner workings of the perceptron to non-specialists, to everyone.
https://marcoheleno.gitlab.io/projects/perceptronexplanator/
