CapoCaccia

Cognitive Neuromorphic Engineering Workshop

User Tools

Site Tools


cc18:relational-networks-for-goal-directed-sensory-motor-task:overview

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
cc18:relational-networks-for-goal-directed-sensory-motor-task:overview [2018/03/28 14:15]
89.206.64.7
cc18:relational-networks-for-goal-directed-sensory-motor-task:overview [2020/01/09 20:31] (current)
Line 35: Line 35:
 functions. functions.
  
-**1.2 Proposed solution**+**2Proposed solution**
  
 We will build a (relatively) simple spiking neuronal network implementing a goal-directed state-to-action We will build a (relatively) simple spiking neuronal network implementing a goal-directed state-to-action
 mapping (SAG) unit. As a core architecture,​ we will use a three-way relation network mapping (SAG) unit. As a core architecture,​ we will use a three-way relation network
 proposed by Peter Diehl [1] (see Fig 1 on the left). This unit consists of four recurrently connected proposed by Peter Diehl [1] (see Fig 1 on the left). This unit consists of four recurrently connected
-populations of neurons (nodes), three of which represent one-dimensional variables and the forth+populations of neurons (nodes), three of which represent one-dimensional variables and the fourth
 one encodes a relation between them (e.g. A + B - C = 0). A crucial step in translating the one encodes a relation between them (e.g. A + B - C = 0). A crucial step in translating the
 proposed architecture in [1] to our specified task (see below) is to change the input from a rate encoded proposed architecture in [1] to our specified task (see below) is to change the input from a rate encoded
Line 62: Line 62:
  
  
-**1.3 Anticipated challenges**+**3Anticipated challenges**
  
 Since in the original model all weights were converging to the desired state by STDP learning, here Since in the original model all weights were converging to the desired state by STDP learning, here
Line 74: Line 74:
 Finally, tuning the populations’ excitation versus inhibition parameters to achieve stable and Finally, tuning the populations’ excitation versus inhibition parameters to achieve stable and
 robust Winner-take-all behavior will also be a challenge. robust Winner-take-all behavior will also be a challenge.
 +
 +**4. Expected working steps**
  
 1. Using Brian2 simulator (and, possibly, ncs_brian library [5]), flash out the connectivity of a 1. Using Brian2 simulator (and, possibly, ncs_brian library [5]), flash out the connectivity of a
Line 89: Line 91:
  
 {{ :​cc18:​relational-networks-for-goal-directed-sensory-motor-task:​sag_unit.png?​nolink&​600 |}} {{ :​cc18:​relational-networks-for-goal-directed-sensory-motor-task:​sag_unit.png?​nolink&​600 |}}
 +Figure 1: LEFT PANEL: Adapted figure from [1] showing a scheme of a three-way relational
 +network. Yellow circles represent populations of LIF-neurons,​ blue circles represent inputs (given
 +two input the network infers the remaining one). Blue arrows depict the direction of connections,​
 +emphasizing recurrent connectivity. For convenience of the "​hard-wired"​ hardware implementation,​
 +the size of the S, A and H populations will be reduced to 256 neurons. The size of the G population
 +will be reduced to 30 neurons. RIGHT PANEL: An example of state-to-action mapping for three
 +various goals: (i) following the stimulus (G1), (ii) avoiding the stimulus (ii), (c) keeping a fixed
 +distance from the stimulus (G3)
 +
  
 **References:​** **References:​**
Line 105: Line 116:
 [4] J. J. Buhmann, Networks of Cooperative Controllers for Distributed and Hierarchical Decision [4] J. J. Buhmann, Networks of Cooperative Controllers for Distributed and Hierarchical Decision
 Making. PhD thesis, ETH Zurich, 2017. Making. PhD thesis, ETH Zurich, 2017.
 +
 +[5] NCS_brian library is available at https://​code.ini.uzh.ch/​ncs/​libs/​ncs_brian
cc18/relational-networks-for-goal-directed-sensory-motor-task/overview.1522246524.txt.gz · Last modified: 2020/01/09 20:31 (external edit)