We propose a new self-organizing hierarchical softmax formulation for neural-network-based language models over large vocabularies. Instead of using a predefined hierarchical structure, our approach is capable of learning word clusters with clear syntactical and semantic meaning during the language model training process. We provide experiments on standard benchmarks for language modeling and s...

The concept of self-organized criticality (SOC) has been evoked by Bak, Tang and Wiesenfeld (BTW) as a possible unifying framework to describe a vast class of dynamically driven systems which evolve spontaneously toward a critical stationary state with no characteristic time or length scale. In this essay we will discuss SOC models that have been studied and to what extent they are related to r...

In the present study we address the question of how persons or families occupy specific locations according to an attractiveness function between persons and flats. We suggest a mathematically formulated model in order to find optimal distributions of persons over flats. The model maximizes the global attractiveness of the ensemble and accounts for various conflicting situations. Its solutions ...

The causes and possible relations between the abundance of fractal structures and 1/f signals found in nature have puzzled scientists for years. Based on observations of computer simulations of the cellular automata sandpile model, the term, self-organized criticality (SOC) was invented by theoretical physicists Bak and coworkers (1987) and was used to explain a universal dynamic organizing pro...