Wagner's SOP model Flashcards
1
Q
phenomena that SOP fixes
A
- one-trial overshadowing
- latent inhibition & its context specificity
2
Q
phenomena that SOP fails at
A
- downshift unblocking
- inhibition
3
Q
hebb & konorski
A
learning-rate parameter: pi
4
Q
SOP activity states
A
- A1 - limited capacity
- A2 - unlimited capacity
limited capacity feature of A1 activity creates overshadowing
5
Q
SOP’s learning rules
A
- rather than on-on –> excitation
- A2 states primed & likened it to the refractory period in nervous transmission
- gives it all RWM features it needs
6
Q
representational activity cycle
self-generated priming
A
an inactive stim’ elements (I) become fully active (A1) when is presented, and fade (A2) eventually returning to I
7
Q
retrieval-generated priming
A
- An associatively-activated representation goes into from inactive but only to its A2 state
- Cannot immediately enter into A1 without passing back through inactive: takes some time; without this feature, model would not work
- At λ the presentation of CS will ‘prime’ most of UCS’s representational elements into their A2 states: generates the CR is Pavlovian conditioning
8
Q
blocking & overshadowing with SOP
A
- during stage 2 - the tone primes many of the shocks elements into their A2 states
- this means that the blocking test stim wont be learned about normally
- this is because the states are in their A1 states while shock’s elements in A2 states
9
Q
SOP explaining latent inhibition with self-generated priming of CS
A
- The “context” is present learning and is subject to the same learning rules as any other stimulus
- During the pre-exposure stage of latent inhibition, the CS is pre-exposed
- This is assumed to allow the context to enter into an association with the CS
- As a result the context will retrieval-generate priming in the CSs representation
- This means that the CS’s elements will be in their A2 states, to some extent; ∴ during Conditioning, they won’t be able to become associated with the UCS
- A confirmed prediction of this is that changing the prediction between Pre-exposure will abolish latent-inhibition, leaving conditioning to work normally