Abstract: A method for visualizing neurons in an Artificial Intelligence (AI) model for autonomous driving. The method includes obtaining, from a number of neurons of the AI model for a task, one or more neurons; determining, for each of the one or more neurons, a respective Region of Interest (ROI) of an input related to the task, wherein the respective ROI is encoded by the one or more neurons for the task; and producing a human-interpretable representation of the determined respective ROI of the input for at least a portion of the one or more neurons, by applying a first operation including Layer-wise Relevance Propagation (LRP).

Abstract: A method for reducing the computational cost for an autonomous driving system is disclosed. The method may include the following steps: a) acquiring data related to a task for operating a vehicle; b) training a deep learning model using the data acquired, wherein the deep learning model includes an encoder and a policy head for the task; c) reducing a complexity of the data acquired in step a) by passing the data to the encoder to produce a compressed latent representation of the data; and d) determining a driving operation by the policy head using the compressed latent representation of the data.

Abstract: A method for driving policy visualization, the method includes (i) receiving, by a processing circuit, perception information that comprises environmental information about an environment of a vehicle and kinematic information regarding a movement of the vehicle; (ii) receiving, by the processing circuit, a multidimensional virtual force field representation of a driving policy applicable to the vehicle; (iii) reducing a dimension of the multidimensional virtual force field representation, based on the received perception information, to produce a reduced dimensional virtual force field representation that conforms with a driving of the vehicle; and (iv) dynamically visualizing, by applying the reduced dimensional virtual force field representation, the driving policy in the driving of the vehicle.

Abstract: A method for explainable representation, the method includes: (a) receiving, by an auxiliary representation network, information regarding an environment of a vehicle; the information being destined to be processed by a policy model, to provide driving related decisions at a current point of time; and (b) generating, by the auxiliary representation network, an interpretable representation of predicted outcomes of the policy model during a period of time that ends after the current point of time.