Skip to content
MARL in Cooperative Environments
Edit this page

Interactive Flashcards

Retrieve the key concepts, equations, and intuitions from memory.

45 cards across five decks. Each deck opens with every card visible, so you can test yourself one card at a time or reveal the whole deck and read it as a revision sheet.

DeckSizeCovers
Background10 cardsPolicies, Joint actions, States and observations, Partial observability, Shared rewards
Coordinate10 cardsJoint actions, Non-stationarity, CTDE, Credit assignment, VDN and QMIX
Communicate10 cardsMessages as actions, Message content, Capacity and noise, Communication cost, Learned protocols
Adapt10 cardsPartner dependence, Training diversity, Agent modelling, Ad hoc teamwork, Cross-play
LLMs as Cooperative Agents5 cardsLLMs as agents, Orchestrated and learned collaboration, Dec-POMDP mapping, Shared rewards and CTDE, Open questions