Short term memory for bipolar temporal patterns

Summary form only given. A study of the short-term memory requirements of temporal pattern recognition prompts the creation of a new model for neural computation. It is hypothesized that neural responses resemble hysteresis loops, instead of the simple sigmoid. The upper and lower halves of the hysteresis loop are described by two equations. Generalizing the two equations to two families of curves accommodates loops of various sizes. It is conjectured that this unit is capable of memorizing the entire history of its inputs.< >