1: \begin{abstract}

2: Artificial spike-based computation, inspired by models of computations

3: in the central nervous system, may present significant performance

4: advantages over traditional methods for specific types of large scale

5: problems. In this paper, we study new models for two common instances

6: of such computation, winner-take-all and coincidence detection. In

7: both cases, very fast convergence is achieved independent of initial

8: conditions, and network complexity is linear in the number of inputs.

9: \end{abstract}

10: