You are here

Somatosensory working memory

Description: 
This model is based on that published in Singh, R., Eliasmith, C. (2006). Higher-dimensional neurons explain the tuning and dynamics of working memory cells. Journal of Neuroscience. 26, 3667--3678. [[pdf](http://compneuro.uwaterloo.ca/cnrglab/?q=system/files/singh.eliasmith.2006.2d+working+mem.pdf)]. It is not the exact published model (contact the authors for the original matlab code), but it functions the same way.
Requirements: 
Peer Reviewed: 
Yes
Publication: 
Singh, R., Eliasmith, C. (2006). Higher-dimensional neurons explain the tuning and dynamics of working memory cells. Journal of Neuroscience. 26, 3667--3678.
Publication URL: 
http://compneuro.uwaterloo.ca/cnrglab/?q=system/files/singh.eliasmith.2006.2d+working+mem.pdf
Instructions: 

1\. To run this demo, open the somatosensory working memory.nef file in Nengo and 'run simulation' by right clicking the network. If that doesn't work, open the script after puting the .layout file in your layouts directory. Click the interactive plots icon to run the model.

2\. Once it's loaded, run it for 3s by right-clicking the **Network Viewer**.

3\. One interesting aspect of this model is that it uses adapting neurons (spike rates change with constant input).

4\. The far more interesting thing about this model is that it explains the data set from Romo et al. (1999) on somatosensory descrimination task in the macaque. The classes of neural response that they indentified is shown in figure 1. No other model has been able to capture all of these response types.

Figure 1 (below): PSTH plots during memorization. The gray bars under the axes indicate the onset of the stimulus, and black bars above the graph mark periods of monotonicity. The higher stimulus
frequency (f1) is marked with darker response curves. a, c, e, Positive monotonic. b, d, f, Negative monotonic. a, b, Early neurons. c, d, Persistent neurons. e, f, Late neurons. [Graph from Romo
et al. (1999).]

4\. Note that the data above is filtered with a Gaussian kernel to make it into smooth firing rates, you will have to compare spike densities. Filtering is not included in this code.

5\. To see different neurons being used, go into the Nengo network viewer, choosed the 2D population and click the inspector (magnifying glass). Go down to 'neurons' and type in some value (less than 200 or it will be slow), and hit enter. This generates a new population of cells (so slightly different tuning curves). Then you can rerun the model.

- Look to see if you can find the classes of neuron experimentally identified above in the spike raster.
- For instance 'c' above has some neurons with a brief initial burst and then reasonably constant firing.
- As well, some neurons have very rapid bursts and then silence ('b'), or prolonged silence until a later increase in firing ('e'). Few neurons slow down firing over time (some lines in 'd').
- To verify that the full set of patterns is present, you have to run the simulation with a variety of inputs, and track a single neuron across different inputs.

Figures: