I'm trying to examine a hypothesis about the statistics of trained "deep" networks. There have been quite a few impressive results published in recent years (most recently, state of the art state detection based on multi-layer neural networks). It's proved to be surprisingly difficult to find code that describes the "final" networks used in these articles.
Does anyone know of a decent source for trained networks that achieved very competitive results? I'm interesting specifically in the layout of trained networks right now, not code that re-produces them.