APRIL 3

Question 1

recap: the promise of computational modeling

Answer 1

1. computational modeling can GROUND affective research in QUANTIFIABLE METRICS that are precisely specified and TESTABLE
2. this can ESTABLISH BIOMARKERS for human research
3. this can BUILD CONNECTIONS between humans and non-human animals for rigorous mechanistic research

Question 2

mood

Answer 2

like emotions, but different

1. LAST LONGER
2. LESS TIGHTLY LINKED TO SPECIFIC EVENTS
3. REFLECT CUMULATIVE IMPACT OF MULTIPLE EVENTS/STIMULI

Question 3

moods can influence how easily…

Answer 3

emotions are elicited

ie. depression increases expression of anger

Question 4

potential adaptive function of moods

Answer 4

increase EFFICIENCY in LEARNING ABOUT OUR ENVIRONMENNT

Question 5

moods interact with outcomes

Answer 5

the outcomes we experience influence our moods (CAUSE of moods)

our moods influence how we experience outcomes (CONSEQUENCE of moods)

Question 6

what is often used in psychology research to influence mood?

Answer 6

monetary outcomes

they can be PRECISELY CONTROLLED along a CONTINUUM

and RELIABLY impact mood

Question 7

ecological momentary assessment (EMA)

Answer 7

used to TRACK SUBJECTIVE WELLBEING in naturalistic settings

repeated data collection to probe Ps affective stat in real time as they go about daily lives

reveals how REAL-WORLD EVENTS (sunshine, sporting event outcomes, visiting new locations) influence mood

Question 8

mood depends on the ______ impact of _______ outcomes

Answer 8

mood depends on the CUMULATIVE impact of UNEXPECTED outcomes

Question 9

variables in the happiness/mood equation

Answer 9

for each trial j up to the current trial t

CR

EV

RPE

w

y

Question 10

CR

Answer 10

value of certain reward, if chosen

Question 11

EV

Answer 11

expected value of the GAMBLE

Question 12

RPE

Answer 12

difference between the ACTUAL OUTCOME and the GAMBLE EV

Question 13

w

Answer 13

weights

capture influence of task variables on momentary happiness

Question 14

y

Answer 14

exponential decay parameter to capture forgetting

Question 15

the same model was fitted to all participants, and the value of the ______ ______ was examined to…

Answer 15

value of the WEIGHT PARAMETERS

was examined to

UNDERSTAND HOW DIFF FACTORS WERE CONTRIBUTING TO HAPPINESS RATINGS

found that RPE weights contributed significantly higher than EV weights

Question 16

model found that which of RPE or EV weights contributed more to happiness?

Answer 16

RPE weights contributed more

so, Ps level of happiness was best explained by model based on RPE and NOT THEIR OVERALL EARNINGs

Question 17

happiness depends not simply on how well things are going, but on…

Answer 17

whether they are GOING BETTER THAN EXPECTED

(likely involves dopamine signalling and BOLD activity in ventral striatum)

Question 18

slot machine experiment setup

Answer 18

Ps played 2 different slot machines with similar levels of PROBABILISTIC REWARD

between games, played Wheel of Fortune and WON or LOST $7 as a MOOD MANIPULATION

Question 19

slot machine experiment results

Answer 19

in those who self-reported HIGH EMOTIONAL INSTABILITY:

^ WINNING INCREASED HAPPINESS and losing decreased happiness

^ associated with changes in BOLD signal in striatum

Question 20

in those who self-reported high emotional instability, what happened to the striatal response to reward following loss?

Answer 20

it was reduced

(evidence of mood manipulation)

Question 21

participants who won Wheel of Fortune chose to replay which game?

Answer 21

game 2

and those who experienced a loss preferred game 1

SUGGESTS that mood manipulation INFLUENCED HOW they experienced the two games that were otherwise the same

Question 22

Ps preferred the game if they encountered it…

Answer 22

in a good mood (ie. after winning $7 in Wheel of Fortune)

and avoided it if in a bad mood (ie. after losing $7 in Wheel of Fortune)

Question 23

what points to individual differences in our sensitivity to the effects of mood?

Answer 23

the Wheel of Fortune win/loss only affect game preference in those who self reported HIGH EMOTIONAL INSTABILITY

Question 24

reinforcement learning

Answer 24

a class of ALGORITHMS that LEARN FROM TRIAL AND ERROR

to PREDICT which STATES of the ENVIRONMENT and which ACTIONS in those states

will MAX reward and MINIMIZE punishment

Question 25

mood's adaptive function

Answer 25

tree lecture (what if fruit increases or decreases in 2 trees at once?) - what if there are INTERDEPENDENT CHANGES BETWEEN STATES? ie. effect of positive mood on learning means that expectations reflect NOT ONLY the reward associated with ONE PARTICULAR STATE (tree) but ALSO the recent OVERALL CHANGES in reward availability in the ENVIRONMENT (all other trees)

Question 26

mood is an efficient way for general environmental factors to...

Answer 26

influence reward expectations without having to DIRECTLY LEARN the impact of these factors on EVERY STATE

Question 27

mood can capture _________ in sources of ______ to increase what?

Answer 27

mood can capture INTERDEPENDENCIES in sources of REWARD to increase LEARNING EFFICIENCY

Question 28

if fruit becomes more abundant in all trees...

Answer 28

an animal foraging will be positively surprised repeatedly, regardless of the tree it visits repeated positive RPE will INDUCE POSITIVE MOOD positive mood will FURTHER INCREASE SUBJECTIVE REWARD of each piece of fruit expectation of reward will increase more rapidly ie. positive surprises will work to increase effect of more positive suprises

Question 29

changes in current reward can predict changes in future reward within a single...

Answer 29

state ie. seasonal changes - springtime

Question 30

positive mood can represent positive _______ in reward availability

Answer 30

momentum this biases perception of rewards (rewards will be perceived as BETTER THAN THEY REALLY ARE, and EXPECTATIONS RAPIDLY CATCH UP with rising rewards)

Question 31

if reward availability is decreasing, what happens to mood and reward perception?

Answer 31

mood decreases because of negative RPEs reward perception decreases because of negative mood (rewards are perceived as less good than they actually are)

Question 32

when is a standard learning algorithm optimal in estimating expected reward?

Answer 32

if reward is determined by the state we're in and if each state is independent

Question 33

Vs

Answer 33

the estimated mean reward of state s in a standard learning algorithm, the estimate is updated at each step by adding a scaled RPE to previous estimate

Question 34

how is the estimate updated in a standard learning algorithm?

Answer 34

by adding a scaled RPE to previous estimate and the RPE is scaled by a learning rate

Question 35

a mood-informed learning algorithm

Answer 35

standard learning algorithm doesn't cut it states are not independent - multiple states are affected by GENERAL ENVIRONMENTAL CHANGES so it's more efficient to UPDATE EXPECTATIONS OF ALL STATES TOGETHER

Question 36

how can we update expectations of all states together?

Answer 36

by keeping track of ALL RECENTLY EXPERIENCED RPEs

Question 37

mt, nt and ft in the mood informed learning algorithm

Answer 37

mt = tracked average of all recent RPEs nt = a learning rate ft = scaling factor

Question 38

ft

Answer 38

degree to which the mood factor influences updating (could be individual diffs in this)

Question 39

mt

Answer 39

tracked average of all recent RPEs

Question 40

nt

Answer 40

learning rate

Question 41

positive and negative moods are adaptive and useful insofar as they persist only...

Answer 41

only until expectations are updated to align with changes in environment

Question 42

what does affective homeostasis rely on?

Answer 42

appropriately updating expectations requires appropriate learning processes

Question 43

failure to appropriately update expectations can lead to what?

Answer 43

maladaptive behaviour ie. failing to learn about negative outcomes could lead to repeated negative surprises (RPEs) that will induce negative mood

Question 44

potential outcome learning mechanism of depression

Answer 44

failure to learn about negative outcomes (overly optimistic expectations) leads to negative RPEs leads to negative mood

Question 45

what does stable mood fluctuation rely on?

Answer 45

BALANCED POS and NEG learning rates these lead to STABLE EXPECTATIONS and so stable fluctuations in mood

Question 46

small imbalances in pos/neg learning rates can lead to...

Answer 46

instability

Question 47

what does a slightly lower negative learning rate lead to?

Answer 47

overly optimistic expectations that are followed by REPEATED and LARGER negative RPEs and PERSISTENT NEGATIVE MOOD

Question 48

if you have neg mood induced by lower neg learning rate and neg RPEs, what can positive surprises lead to?

Answer 48

an ESCALATING FEEDBACK LOOP positive surprises = postive RPEs will change expectations and improve mood lead to overly optimistic expectations and then mood stabilizes and expectations catch up and outcomes FALL SHORT of expectations followed by negative surprises and neg mood

Question 49

hypothetical mechanism of mood instability in bipolar

Answer 49

slightly lower negative learning rate leads to negative surprises > lower mood > biased perception bias perception = overly low expectations overly low expectations leads to positive suprises positive surprises > higher mood > biased perception feedback loop

Question 50

what's a key implication of learning rates and effects on mood

Answer 50

depending on how we learn about our environments and actions we can have the SAME OBJECTIVE EXPERIENCE yet very DIFFERENT SUBJECTIVE EXPERIENCES

Question 51

conclusions

Answer 51

1. mood may serve as a REP of the MOMENTUM of CHANGES IN REWARD 2. this momentum can be used to ADJUST LEARNING to INCREASE EFFICIENCY when there are DEPENDENCIES in changes across time 3. could have been EVOLUTIONARILY ADAPTIVE in supporting rapid adaptation in changing environments 4. understanding the proper function of moods can help generate hypotheses about mood dysfunction in disorders like ANXIETY and DEPRESSION