My specific report is essentially a long winded explanation of top down processing that exists in hippocampal place cells. I only read several pages of the article so to exemplify top down processing. I never heard of attractor dynamics before reading this article. I thought the vocabulary was odd in this article. I think this is because the article comes from the United Kingdom. Still, the information parallels what I am learning in school.
The full article I used can be found here: http://www.hindawi.com/journals/np/2011/182602/
Place Cells, Discrete Attractor Dynamics, and Attractor Landscapes

We all intake external stimuli through the senses and through transduction we are able to perceive a representation of that stimuli. Some stimuli include objects such as trees, walls, or even fields. Place cells are thought to create cognitive maps which help us perceive ourselves in the world. Place cell firing occurs in the hippocampus whenever we enter a particular environment. Place cells react to the changes in the environment and memories of past environments are used to help us perceive the world.
Place cells help human beings form cognitive maps of the world before them. Humans create this cognitive map when place cells, located in the hippocampus, fire. Changes in place cell firing patterns are made when the environment changes. For example, when a person walks from a field into a house, the pattern of place cell firing in the hippocampus will alter. This alteration in place cell firing is called remapping.
Whenever the place cells remap, the ultimate goal is a stable state made up of many neurons sharing synaptic gaps. Attractor networks are used to explain this ultimate goal. Attractor networks are networks of neurons that develop so to have a stable state over time. An attractor landscape is previously known knowledge about the environment. There are two types of attractor dynamics: discrete attractor dynamics and continuous attractor dynamics. This paper will concern itself with discrete attractor dynamics.
Discrete attractor dynamics indicates when place cells resist small changes in environment, but respond collectively to large, sudden, or abrupt, changes (e.g. if you are a performer, the curtain is drawn and you are presented with a giant theater).
Whenever humans are presented with a stimulus the attractor network (network of neurons gravitating towards stability) compares incoming information with an attractor landscape. Attractor landscapes contain knowledge that was known prior to the stimulus. The incoming information is compared to the attractor landscape and from this comparison one of two states is formed. The first state is pattern completion, which pertains to minor changes (less stimulation). The second state is pattern separation, which pertains to major changes in the environment (abrupt). That is, whenever we receive sensory input, our place cells either remap or stay the same based on how our prior memory compares to the stimulus environment.
It is interesting to note that Wills and colleagues (2005) showed that place cell activity remained stable whenever incremental changes were made to a circular or square environment. Only when cumulative changes became great enough did the place cells remap. In this case, whenever the incremental changes failed to produce the remapping of place cells, a state of completion is said to have been made. In contrast, whenever the cumulative changes of the environment were great enough, the attractor network remapped and thus achieved a state of pattern separation (Wills et al., 2005). Furthermore, the entire time that incremental changes were being made, it should be noted that previous attractor landscapes (prior knowledge related to the stimuli) were constantly being compared to the incoming stimuli. Conversely, In another study, Leutgeb and colleagues (2005) recorded place cell activity during incremental changes in environment activity and found that the patterns did not jump from one pattern to the next. Rather, the place cells gradually changed (Leutgeb et al., 2005). Differences in the two studies have been explained by the attractor landscape, which is a representation of prior knowledge meant to help us perceive the world. Attractor landscapes provide us with a framework, and from that framework our place cells shift or stay the same. The studies suggest that attractor landscapes vary from person to person.
Place cells remap into arrangements determined by discrete attractor networks states, which are influenced, by the attractor landscapes (prior knowledge) and incoming stimuli. Whenever we see the world, our prior knowledge of the world combines with the incoming stimuli and we create a cognitive map. Two people may have different amounts of prior knowledge about a certain place in the world and so their remapping thresholds will differ and thus their place cells will remap differently.
References
Jeffery. K J. (2011). Place Cells, Grid Cells, Attractors, and Remaping. Neural Plasticity, vol. 2011, Article ID 182602, 11 pages.
Leutgeb, J. K., Leutgeb S., Treves A. et al. (2005) Progressive transformation of hippocampal neuronal representations in “morphed” environments. Neuron, vol. 48, no. 2, pp. 345–348.
Wills, T. J., Lever, C., Cacucci, F., Burgess, N., & O’Keefe J. (2005). Attractor dynamics in the hippocampal representation of the local environment. Science, vol. 308, no. 5723, pp. 873–876.
http://en.wikipedia.org/wiki/HippocampusQuote:
Video 1.