This paper focuses on a specific aspect of human visual discrimination from computationally generated solutions for CAAD ends. The bottleneck at work here concern informational ratios of discriminative rates over generative ones. The amount of information that can be brought to a particular sensory modality for human perception is subject to bandwidth and dimensional limitations. This problem is well known in Brain-Computer Interfaces, where the flow of relevant information must be maintained through such interaction for applicative ends and adoption of use in many fields of human activity. While architectural modeling conveys a high level of complexity in its processes, let alone in the presentation of its generated design solutions, promises in applicative potentials of such interfaces must be made aware of these fundamental issues and need developments of appropriate sophistication. This paper addresses this informational bottleneck by introducing a method to retrieve spatial information from the rapid serial visual presentation of generated pictures. This method will be explained and defined as inverse modeling, based on inverse graphics, and its relation to human visual processing.