6 Discussion and Future Work

  The results presented in the previous chapter demonstrate rigorously that activity-dependent plastic lateral interactions could be responsible for the tilt aftereffect. The same self-organizing principles that result in sparse coding and reduce redundant activation may also be operating over short time intervals in the adult, with quantifiable psychological consequences. This finding demonstrates a potentially important computational link between development, structure, and function.

To clarify exactly which aspects of human behavior have been replicated in the model, the following section will review the available psychophysical evidence and compare it to the results found in the model. The model accounts for the great majority of such evidence to date, but it does not yet demonstrate the saturation effect or decay of the TAE in darkness. Later sections will discuss how the cortex may implement the functions performed by the model, using the available biophysical evidence as constraints on the types of models considered to be plausible. The adult cortex might implement quite different mechanisms for plasticity than are present in the developing animal. Based on the results in chapter 5, a number of predictions for verification of the model by human and animal experiments are also discussed. For instance, one could measuring cortical activity analogously to the activity plots in the previous chapter. This thesis also opens up a variety of new directions for modeling function with RF-LISSOM, some of which are proposed at the end of this chapter.