lecture 9- dopamine, reward & reinforcement Flashcards

You may prefer our related Brainscape-certified flashcards:
1
Q

what is dopamine?

A

a type of neurotransmitter

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

what is the importance of dopamine?

A

important for learning, cognition and behavior.

important for coding prediction errors which drives our learning as we only learn things when we are surprised.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

what is a negative reinforcer?

A

decrease the frequency/probability of a certain action- avoiding doing the action, not a punishment just avoidance.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

what is an example of negative reinforcement?

A

you tidy up the house to avoid getting shouted at by housemate.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

what is a prediction error?

A

errors in the prediction of rewards signal the inappropriate/appropriate nature of the actions performed to obtain them.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

give the 2 examples of Reward Prediction Errors (RPEs)

A
  1. unexpected rewards

2. failure to obtain a predicted reward

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

why are RPEs important?

A

outcomes are unexpected and therefore are important in driving our learning.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

explain the Rescorla-Wagner model.

A

basically explains the various forms of classical conditioning reliably.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

why would it be good to find neurons that respond to reward stimuli?

A

if we find neurons that encode the unpredictability of the reward it would be very beneficial for learning.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

dopaminergic neurons have what receptors? and what is the release of dopamine associated with?

A

dopamine receptors, and it is associated with pleasure.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

what pathways in the brain are associated with dopamine?

A
  1. Mesostriatial- basal ganglia which influences motor control.
  2. mesolimbocortical pathway -including hippocampus, which influences memory and reward processing.

(most are middle/front of brain effected by dopamine).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

where contains a large proportion of dopamine neurons?

A

midbrain- Ventral tegmental area (VTA) .

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

where do dopamine neurons project to?

A

Striatum and anterior cingulate cortex.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

what do dopamine neurons respond to?

A

presentation of rewards.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

an example of a prediction error is with monkeys and apple juice. explain the levels of dopamine throughout this experiment.

A

before the task- dopamine neurons respond to unpredicted rewards.
during learning- reward became increasingly predictable.
when the reward becomes more predictable less dopamine is fired.
if the monkey doesnt get juice- no dopamine is fired- predictable rewards elicit no change.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

what happens if rewards are delivered at the wrong time ?

A

there is a decreased activity at the time that rewards are expected, followed by increased activity at the time of delivery. (decreased dopamine when reward not given when they think they are getting a reward).

17
Q

what do Schultz et al (1997) suggest about reward prediction?
(in monkeys remember)

A

during learning the dopamine response switched from the reward (US) to the reward predictor (CS). rewarding experienced at time of Que not actually when given reward.
neurons also involved in reward prediction - Que itself could be like a reward itself.

18
Q

VTA in humans - D’ Ardenne et al (2008) experiment explain…

A

P’s played a card guessing game - guess whether the number on the right would be bigger/less than on the left- they won $1 if guessed correctly and lost $1 if guessed incorrectly.
results= activity in the VTA increased when they UNEXPECTEDLY got the $1.. and was greater as the probability of the reward increased.
therefore some evidence for prediction errors in humans however still debated (Duzel et al, 2009) as there was NO response for negative prediction errors.

19
Q

what is the equation for the dopamine response??

A

dopamine response= actual reward- predicted reward.

following Rescorla- Wagners learning principle.

20
Q

what does Matsumoto et al (2013) suggest about VTA dopamine neurons signalling cognitive motivation?

A

experiment in monkeys:performing a delayed matching-to-sample task that required working memory and visual search.
results= dopamine neurons responded to task events associated with cognitive operations.
The response of dopamine neurons in the ventral tegmental area is modulated by task difficulty!!

21
Q

what did Walton et al (2007) show?

A

dopamine neurons coding how much effort is needed to put in for output.
Coding for = Reward - Amount of effort

22
Q

where are ‘reward expectation’ neurons found in monkeys and rats brains?

A

the Striatum.

(part of the basal ganglia) also known as the ICE CREAM CONE.- the nucleus accumbens

23
Q

what two parts in the Stratium composed of?

A
  1. The Caudate

2. The Putamen

24
Q

what did Pessiglione et al (2006) show?

A

evidence for prediction errors in humans by looking at the response in the stratium from fMRI.
L-DOPA (enhances dopamine) and haloperidol (reduces dopamine) was administered during instrumental learning.
L-DOPA= ehanced the RPE signal and choose the most rewarding action relative to subjects treated with haloperidol.

25
Q

where else do dopamine neurons target in the brain??

A

Cingulate cortex- anterior and mid portions
respond to rewarding stimuli also.
Anterior Cingulate Cortex (ACC).

26
Q

what did Kennerley et al (2011) find in monkeys ACC when making choices between rewards?

A

they found neurons that responded to +ve, and -ve prediction errors.
in addition to neurons responding to both at the same time.

27
Q

Ribas-Fernandes (2011) examined RPEs in HUMAN ACC’s what did he find?

A

after examining human fMRI scans after performing a task where there is a sub goal leading to an overall outcome, they found activity in the ACC that signalled prediction errors for the individual actions leading to a goal.

28
Q

what happened to the rats that were hooked up to a self administration of dopamine?

A

if the rat pressed the lever more dopamine was released, rats ended up passing out because they repeatedly released the dopamine for pleasure.

29
Q

how does cocaine work on dopamine?

A

directly on dopamine neurons by blocking dopamine re-uptake. increasing the concentration of dopamine at the synapse - sense of euphoria.
(showing dopamine is rewarding).
cocaine effects transmission- dopamine cant go back to be recycled in the pre synaptic membrane.

30
Q

cocaine and temporal coding: what happens when rats are assked to wait for a reward when on drugs?

A

cocaine rats cannot wait long enough, they are too motivated for getting the reward + are expecting the reward so much they cannot wait to press the lever.

31
Q

summary - the Mesolimbic Dopamine system what are the 4 main points to this?

A
  1. learns to predict rewards following PE’s.
  2. signals the anticipation of an upcoming reward.
  3. motivates the performance of action for getting rewarding outcomes.
  4. disturbed by drugs- which change motivations.
32
Q

what does Behrens et al (2008) suggest about PE’s?

A

learning about others may be underpinned by similar reinforcement learning processes.
we make predictions what other peoples actions may be in a similar way to how we make predictions about ourselves.
therefore we can correct predictions about others when the prediction was in error.

33
Q

discuss Behrens et al (2008) experiment on PE’s beyond reward learning looking at social cognition…

A

looked at humans fMRI, whilst they performed a decision making task which required them to see which is the best out of two options and were rewarded.

  • on each trial they recieved advice outside of the scanner.
  • subjects used the probability of a reward being received and the probability of the advice not being true (social learning).
  • results found: activity in the nucleus accumbens/ ventral stratium signalled the RPEs for the subjects own learning about rewards.
34
Q

Kennerley, Behrens & Wallis (2011) found a double dissociation between which two types of neurons?

A

Orbitofrontal neurons and anterior cingulate cortex neurons.