Section 1 of 1
Day 17 · Tue 27 Oct · Agent Foundations: Daily Quiz
Day 17 · Tue 27 Oct · Agent Foundations (D.4) · Aden Power
This is a simple quiz to help you recall what you have learned today, and to confirm you have achieved the learning outcomes. The questions are deliberately easy. Passing this quiz is *not* a replacement for being in the course. You do not need to submit the quiz, and we will not grade it.
Nothing is collected about who you are. If you do submit, your score and the correct answers appear immediately.
This is a simple quiz to help you recall what you have learned today, and to confirm you have achieved the learning outcomes. The questions are deliberately easy. Passing this quiz is *not* a replacement for being in the course. You do not need to submit the quiz, and we will not grade it.
Nothing is collected about who you are. If you do submit, your score and the correct answers appear immediately.
What is the core contrast an “embedded agent” is defined against?
What is the core contrast an “embedded agent” is defined against?
Question Type
Short answer
Paragraph
Multiple choice
Checkboxes
Dropdown
File upload
Linear scale
Rating
Multiple choice grid
Checkbox grid
Date
Time
A supervised learner versus a reinforcement learner
A dualistic or Cartesian agent (like AIXI) that sits outside its environment and interacts through clean input/output channels
A large model versus a small model
An agent with memory versus an agent without memory
Add option
or
add "Other"
…
Answer key
(1 point)
Loading...
Demski and Garrabrant’s “Embedded Agency” splits the problem into four parts. Which of the following are those four? (Select all that apply.)
Demski and Garrabrant’s “Embedded Agency” splits the problem into four parts. Which of the following are those four? (Select all that apply.)
Question Type
Short answer
Paragraph
Multiple choice
Checkboxes
Dropdown
File upload
Linear scale
Rating
Multiple choice grid
Checkbox grid
Date
Time
Embedded world-models
Decision theory
Robust delegation
Subsystem alignment
Reward shaping
Scalable pretraining
Add option
or
add "Other"
…
Answer key
(3 points)
Loading...
In Newcomb’s problem, what do the two classic decision theories recommend?
In Newcomb’s problem, what do the two classic decision theories recommend?
Question Type
Short answer
Paragraph
Multiple choice
Checkboxes
Dropdown
File upload
Linear scale
Rating
Multiple choice grid
Checkbox grid
Date
Time
Both CDT and EDT one-box
Both CDT and EDT two-box
CDT two-boxes; EDT one-boxes
CDT one-boxes; EDT two-boxes
Add option
or
add "Other"
…
Answer key
(1 point)
Loading...
What is the central move made by Functional Decision Theory?
What is the central move made by Functional Decision Theory?
Question Type
Short answer
Paragraph
Multiple choice
Checkboxes
Dropdown
File upload
Linear scale
Rating
Multiple choice grid
Checkbox grid
Date
Time
It maximises expected reward over the next time step only
It refuses to reason about predictors at all
It treats the choice as selecting the output of a decision algorithm, so every instance and every accurate prediction of that algorithm returns the same answer
It always defects in prisoner’s dilemmas
Add option
or
add "Other"
…
Answer key
(1 point)
Loading...
Name one way today’s deployed AI systems are already embedded rather than Cartesian - they affect their own training data, run copies of themselves, are modified by their operators. Which of the four subproblems does that make most urgent?
Name one way today’s deployed AI systems are already embedded rather than Cartesian - they affect their own training data, run copies of themselves, are modified by their operators. Which of the four subproblems does that make most urgent?
Question Type
Short answer
Paragraph
Multiple choice
Checkboxes
Dropdown
File upload
Linear scale
Rating
Multiple choice grid
Checkbox grid
Date
Time
Answer key
(0 points)
Loading...
Link to Sheets
Summary
Question
Individual
Insights
Total points distribution
Loading...
Settings
Release grades
Respondent settings
Global quiz defaults
Default question point value
Point values for every new question
Responses
Manage how responses are collected and protected
Presentation
Manage how the form and responses are presented
Defaults
Form defaults
Settings applied to this form and new forms
Question defaults
Settings applied to all new questions