Thus, for example, if a scene having an animal, such as a dog, is presented or sensed at the input of an HTM, the HTM, having previously learned what a dog looks like, will output some distribution of likely causes of that input, with the ???dog??? cause likely having the highest probability.