3. Pavlovian Learning Part 2

CS

Conditioned Stimulus

US

Unconditioned Stimulus

UR

Unconditioned Response

CR

Conditioned Response

Compensatory Conditional Response

A conditioned response that compensates for a disruption in homeostasis.

Smartphones Example

CS: Sound - US: Message from your 'crush' → UR: heightened emotion.

Restaurants Example

CS: Smell - US: Delicious food → UR: Mouth watering.

Dogs Example

CS: Dog - US: Growling & Lunging → UR: Fear response.

Test Anxiety Example

CS: Test - US: Failing test grade → UR: Anxiety.

Heroin Use First Time

1st time: Euphoria and departure from homeostasis.

Heroin Use Fifth Time

5th time: Euphoria and departure from homeostasis.

Heroin Use Tenth Time

10th: Euphoria and departure from homeostasis.

Heroin Use Twentieth Time

20th: Euphoria and departure from homeostasis.

Heroin Use Fiftieth Time and how you feel?

5th: You feel NORMAL.

Function of the Conditioned Response

The CR prepares the organism for the US.

Extinction

The procedure of repeatedly presenting the CS without the US.

Spontaneous Recovery

The sudden reappearance of a behavior following its extinction.

Extinction Burst

A temporary increase in behavior before it decreases during extinction.

New Context

A situation where the conditioned response may not occur due to lack of associated stimuli.

Stimulus Substitution Theory

Pavlov originally theorized that the CS substitutes for the US. This theory assumes that the CR is essentially the same as the UR.

Preparatory Response Theory

This is the idea that the CR prepares the organism for the appearance of the US.

Compensatory Response Theory

One variation of preparatory response theory that suggests the CR prepares the organism for the US by compensating for its effects.

Rescorla-Wagner Model

Attempts to account for conditioning on a trial-by-trial basis. Assumes that a limited amount of conditioning can occur in the pairing of 2 stimuli and is determined by the characteristics of the US, the maximal value of the UR, and the extent of prior learning.

Conditioned Response (CR)

The response that is learned through conditioning, which is triggered by the conditioned stimulus (CS).

Unconditioned Response (UR)

The natural response that occurs without conditioning, triggered by the unconditioned stimulus (US).

Conditioned Stimulus (CS)

A previously neutral stimulus that, after being paired with an unconditioned stimulus, begins to trigger a conditioned response.

Unconditioned Stimulus (US)

A stimulus that naturally and automatically triggers a response without prior conditioning.

Higher-Order Conditioning

A form of conditioning where a conditioned stimulus is paired with a new stimulus, leading to the new stimulus eliciting the conditioned response.

Elicitation of the CR

The process by which the conditioned response is triggered by the conditioned stimulus.

Qualitative differences between CRs and URs

Differences in the nature of the responses, such as the actual molecular makeup of the saliva.

Conditioned Compensatory Responses

Responses that are the opposite of the conditioned response, often observed in situations like drug overdoses.

Acquisition of the CR is described by what equation?

Described by the equation: ΔVn = c(λ - Vn - 1).

ΔVn

Represents the change in associative strength on trial n.

c

A constant that determines the rate of conditioning.

λ

Total amount of learning/ conditioning possible

Vn

The associative strength prior to this trail for this trial .

Limitations of Stimulus Substitution Theory

The CR and UR are not the same; the CR tends to be weaker and less reliable than the UR.

CR may appear slower than UR, why?

The conditioned response can take longer to manifest compared to the unconditioned response.

Repeated pairing of CS and US

Produces a connection between the CS and US areas in the brain, leading to the CR.

Rescorla-Wagner equation

ΔVn = c(λ - Vn - 1) describes the acquisition of the conditioned response. RWM

D

Delta or Change

V

Strength of the Association

n

specific trail for which learning is being measured

lambda

the total amount of learning that can occur (this # is arbitrary but is typically set at 100)

Vn - 1

the Strength of the Association that was established on the previous trial

Learning Prediction Formula

ΔVn = c(λ - Vn - 1)

Example of Learning Prediction through pairing of stimulis

Suppose we pair the smell of vanilla with eating a fresh-baked chocolate chip cookie.

Initial Value of V

Initially V starts at 0 before there are any pairings because no learning has occurred.

Salience Constant Example

Let's assume the vanilla isn't very salient (c = 0), but the cookies are pretty salient (c =0.5), so let's set c at 0.25.

ΔV1 Calculation

ΔV1 = 0.25(100 - 0) = 25; 25% of all the learning will occur on the first trial

ΔV2 Calculation

ΔV2 = 0.25(100 - 25) = 18.75; 25 + 18.75 = 43.75% of the learning has occurred

ΔV3 Calculation

ΔV3 = 0.25(100 - 43.75) = 14.06; 43.75 + 14.06 = 57.81%

ΔV4 Calculation

ΔV4 = 0.25(100 - 57.81) = 10.55; 57.81 + 10.55 = 68.36%

ΔV5 Calculation

ΔV5 = 0.25(100 - 68.36) = 7.91; 68.36 + 7.91 = 76.27%

ΔV6 Calculation

ΔV6 = 0.25(100 - 76.27) = 5.93; 76.27 + 5.93 = 82.20%

ΔV7 Calculation

ΔV7 = 0.25(100 - 82.20) = 4.45; 82.20 + 4.45 = 86.65%

ΔV8 Calculation

ΔV8 = 0.25(100 - 86.65) = 3.34; 86.65 + 3.34 = 89.99%

ΔV9 Calculation

ΔV9 = 0.25(100 - 89.99) = 2.50; 89.99 + 2.50 = 92.49%

ΔV10 Calculation

ΔV10 = 0.25(100 - 92.49) = 1.88; 92.49 + 1.88 = 94.37%

ΔV11 Calculation

ΔV11 = 0.25(100 - 94.37) = 1.41; 94.37 + 1.41 = 95.78% of the learning has occurred

CR Prediction

The Rescorla-Wagner model predicts that it will take 11 trials to elicit a CR that is 95% of the UR

Altering Salience Example

CS = 1 and US = 1; Maybe a really noxious odor and an intense electric shock...

CS and US Example

CS: Ammonia; US: 200 mV shock

First Trial Learning with High Salience

ΔV1 = 1(100 - 0) = 100; 100% of all the learning will occur on the first trial

RW Model in Neuroscience

At present, the RW model is being extensively used in Neuroscience!

Dopamine Circuits

Learning generally results in alteration of Dopamine circuits in the striatum (the part of your brain that is sensitive to Reward)

Reward Prediction Errors (RPEs)

How your cells respond to the presentation, or lack thereof, of a US following the presentation of a CS produces interesting patterns of responses referred to as Reward Prediction Errors (RPEs)

Dopamine Neurons Activity

Most dopamine neurons signal a reward prediction error: Activated by more reward than predicted (positive prediction error), Remain at baseline activity for fully predicted rewards (zero prediction error), Show depressed activity with less reward than predicted (negative prediction error)

Impact of Drugs on Dopamine

Drugs can generate, hijack, and amplify the dopamine reward signal and induce exaggerated, uncontrolled dopaminergic effects...