COG200 UPDATED EXAM SCRIPT QUESTIONS AND
ANSWERS FIRM A+
✔✔TRUE or FALSE:
Pascal's Expected Objective Value always wants to maximize expected objective value.
- ✔✔TRUE
✔✔Subjective value considers _______, unlike objective value. - ✔✔internal value
✔✔Global workspace = _________ - ✔✔Working Memory
✔✔This explains how modules communicate: - ✔✔Cognitive Architecture
✔✔What are the 3 Levels of Marrs Reinforcement Learning, and their explanations? -
✔✔Functional: Specifying the problem.
Algorithmic: Enabling the problem to be solved.
Physical: Solving the problem.
✔✔What is the policy? - ✔✔Mapping from a particular state to what action should be
taken. (state to action)
✔✔What is the reward? - ✔✔Mapping from a particular state to some quantity/internal
currency. (state to quantity)
✔✔What is the value? - ✔✔Expected future reward - leads to the reward. (state to
future reward)
✔✔Learning Rate (alpha) in reinforcement learning refers to the ________. - ✔✔Update
parameter.
✔✔Discount Factor (gamma)in reinforcement learning refers to ________. - ✔✔The
importance of future reward relative to immediate reward, represented on a scale from 0
to 1. 0 being only caring for only immediate, 1 being only caring for only future.
✔✔Eckman is known for his studies on the innateness of _______. - ✔✔Emotions
✔✔TRUE or FALSE
Emotions are affect programs that coordinate multiple aspects of emotions. - ✔✔TRUE
, ✔✔Reinforcement Learning selects actions that maximize __________. - ✔✔Long Term
expected utility.
✔✔What type of inference is this?
Probabilistic prior knowledge/beliefs + new date = probabilistic updated
knowledge/beliefs. - ✔✔Bayesian Inference
✔✔What are examples of underspecified problems? i.e. Poverty of the Stimulus. -
✔✔Perception, Language, and Infant Cognition.
✔✔What is Innateness (theoretically defined)? - ✔✔Present at birth, product of internal
causes, high heritability, reliably develops across all normal environments.
✔✔What are Tacit Knowledge/Principles? - ✔✔Information within a system that the
person cannot readily articulate.
✔✔TRUE or FALSE:
Kahneman believes that fast thinking consists of heuristics and is error-prone. (rule of
thumb) - ✔✔TRUE
✔✔What is Greene's view on morality judgments? - ✔✔Intuitive moral responses are
based on emotion and are error prone.
✔✔What is Mikhail's view on morality judgments? - ✔✔Intuitive moral responses reflect
a sophisticated moral grammar.
✔✔What is Subjective Internal Value (Subjective Utility)? - ✔✔An internal common
currency in the mind and brain.
✔✔What defines a modular system? - ✔✔Mandatory, fast, domain-specific,
encapsulated, inaccessible, and discretely localized.
✔✔TRUE or FALSE
Subcortical is fast and automated - ✔✔TRUE
✔✔What is moral dumbfounding? - ✔✔People possess strongly ended moral judgments
without any rational justification.
✔✔What is expected consequentialism? - ✔✔An action is morally wrong if the agent
expects that the action increases the chances of harm, relative to alternative actions
available to the agent.
ANSWERS FIRM A+
✔✔TRUE or FALSE:
Pascal's Expected Objective Value always wants to maximize expected objective value.
- ✔✔TRUE
✔✔Subjective value considers _______, unlike objective value. - ✔✔internal value
✔✔Global workspace = _________ - ✔✔Working Memory
✔✔This explains how modules communicate: - ✔✔Cognitive Architecture
✔✔What are the 3 Levels of Marrs Reinforcement Learning, and their explanations? -
✔✔Functional: Specifying the problem.
Algorithmic: Enabling the problem to be solved.
Physical: Solving the problem.
✔✔What is the policy? - ✔✔Mapping from a particular state to what action should be
taken. (state to action)
✔✔What is the reward? - ✔✔Mapping from a particular state to some quantity/internal
currency. (state to quantity)
✔✔What is the value? - ✔✔Expected future reward - leads to the reward. (state to
future reward)
✔✔Learning Rate (alpha) in reinforcement learning refers to the ________. - ✔✔Update
parameter.
✔✔Discount Factor (gamma)in reinforcement learning refers to ________. - ✔✔The
importance of future reward relative to immediate reward, represented on a scale from 0
to 1. 0 being only caring for only immediate, 1 being only caring for only future.
✔✔Eckman is known for his studies on the innateness of _______. - ✔✔Emotions
✔✔TRUE or FALSE
Emotions are affect programs that coordinate multiple aspects of emotions. - ✔✔TRUE
, ✔✔Reinforcement Learning selects actions that maximize __________. - ✔✔Long Term
expected utility.
✔✔What type of inference is this?
Probabilistic prior knowledge/beliefs + new date = probabilistic updated
knowledge/beliefs. - ✔✔Bayesian Inference
✔✔What are examples of underspecified problems? i.e. Poverty of the Stimulus. -
✔✔Perception, Language, and Infant Cognition.
✔✔What is Innateness (theoretically defined)? - ✔✔Present at birth, product of internal
causes, high heritability, reliably develops across all normal environments.
✔✔What are Tacit Knowledge/Principles? - ✔✔Information within a system that the
person cannot readily articulate.
✔✔TRUE or FALSE:
Kahneman believes that fast thinking consists of heuristics and is error-prone. (rule of
thumb) - ✔✔TRUE
✔✔What is Greene's view on morality judgments? - ✔✔Intuitive moral responses are
based on emotion and are error prone.
✔✔What is Mikhail's view on morality judgments? - ✔✔Intuitive moral responses reflect
a sophisticated moral grammar.
✔✔What is Subjective Internal Value (Subjective Utility)? - ✔✔An internal common
currency in the mind and brain.
✔✔What defines a modular system? - ✔✔Mandatory, fast, domain-specific,
encapsulated, inaccessible, and discretely localized.
✔✔TRUE or FALSE
Subcortical is fast and automated - ✔✔TRUE
✔✔What is moral dumbfounding? - ✔✔People possess strongly ended moral judgments
without any rational justification.
✔✔What is expected consequentialism? - ✔✔An action is morally wrong if the agent
expects that the action increases the chances of harm, relative to alternative actions
available to the agent.
