Day 6: Introduction to Representations & Neural Networks

•          Introduction to Representations

•          A Computational Model of a Neuron

•          Local Coding & Vector Coding

•          Examples of Feed Forward Networks

•          Mathematical Analyses of Performance

•          Representations in Neural Networks

•          Start reading Paul Churchland’s “What Happens to Reliabilism When it is Liberated from Propositional Attitudes?” (on ereserve)

 

Introduction to Representations

•          What are representations? What are some of the different approaches to a theory of representation?

•          Why postulate them as part of a theory of mind or knowledge?

•          What might some constraints be on a theory of representation?

 

A Computational Model of a Neuron

•          What are the parts of a Neuron?

•          What do synapses do?

•          What does the soma do?

•          Axons and dendrites?

•          A computational version of the above

•          To what extent is this biologically realistic?

 

Local Coding & Vector Coding

•          Local Coding and Vector Coding are sometimes understood as types of representations

•          What is local coding?

•          What is vector coding?

 

Examples of Feed Forward Networks

•          NetTalk

•          Cottrell’s Face Net

•          EMPATH

•          How are they trained?

 

Mathematical Analyses of Performance

•          State space analysis of hidden units

–        Hierarchical cluster plotting

•          State space analysis of synapses

•          Representations in Neural Networks

•          What, if anything, do patterns of activation across neurons represent?

•          What, if anything, do synapses represent?