A reinforcement learning agent learns mainly from what?
AnswerRewards and penalties for its actions
In reinforcement learning, an agent tries actions and learns which ones earn the most reward over time.
AI · Facts & stats
Agents facts & stats
7 fact-checked facts about AI agents, each with the reason behind it. 5 more are in today's round and join this page after it closes.
The basics
Who proposed the "imitation game" in 1950 to ask whether machines can think?
AnswerAlan Turing
Alan Turing described the imitation game, now called the Turing test, in his 1950 paper.
Going deeper
The term "artificial intelligence" was coined for a 1956 workshop at which college?
AnswerDartmouth College
The 1956 Dartmouth Summer Research Project on Artificial Intelligence is seen as the founding of the field.
In reinforcement learning, what is an agent's rule for choosing actions called?
AnswerIts policy
A policy maps situations to actions, and learning aims to find a policy that earns high reward.
What is a multi-agent system?
AnswerSeveral agents interacting in a shared environment
Multi-agent systems study how many agents cooperate or compete in the same environment.
Expert level
In the PEAS framework for describing agents, what does the "A" stand for?
AnswerActuators
PEAS means Performance measure, Environment, Actuators and Sensors; actuators are how an agent acts.
In reinforcement learning, what does the discount factor (gamma) control?
AnswerHow much future rewards count versus immediate ones
A discount factor between 0 and 1 shrinks the value of rewards that arrive further in the future.
Know these cold?Today's round is five questions like these, 20 seconds each. Your streak and level follow you to every arena.Play today's round