Experimental observations from neuroscience have suggested that the cognitive process of human brain is realized as probabilistic reasoning and further modelled as Bayesian inference. However, it remains unclear how Bayesian inference could be implemented by network of neurons in the brain. Here a novel implementation of neural circuit, named sampling-tree model, is proposed to fulfill this aim. By using a deep tree structure to implement sampling with simple and stackable basic neural network motifs for any given Bayesian networks, one can perform local inference while guaranteeing the accuracy of global inference. We show that these task-independent motifs can be used in parallel for fast inference without intensive iteration and scale-limitation. As a result, this model utilizes the structure benefit of neuronal system, i.e., neuronal abundance and multihierarchy, to perform fast inference in an extendable way. Index Terms-Sampling-tree model, neural network, Bayesian inference, importance sampling, probabilistic population coding I. INTRODUCTION U NDERSTANDING how the brain works is one of the most challenging problems in 21 century. Our brain can represent probability distribution [1], [2], [3]. The cognitive and perceptive process of the brain is a process of probabilistic reasoning, which has been indicated by a number of psychological and neuroscience experiments [4], [5]. From the macroscopic level, Bayesian models have shown their ability of explaining how the brain perceives the world and have been successfully used in various fields of brain