Blame attribution isoneofthekeyaspects ofaccountable decision making, asit provides means to quantify the responsibility of an agent for a decision making outcome. Inthis paper,we study blame attribution inthe contextof cooperative multi-agent sequential decision making.

We consider a general sum partially observableMarkovgamewhere agents ofdifferent types share asingle policy network, conditioned on agent-specific information. This paper aims at i) formally understanding equilibria reached by such agents, and ii) matching emergent phenomena ofsuch equilibria toreal-worldtargets. Parameter sharing with decentralized execution has been introduced as an efficient way to train multiple agents using a single policy network.

Planning enables autonomous agents to solve complex decision-making problems by evaluating predictions of the future. However, classical planning algorithms often become infeasible in real-world settings where state spaces are high-dimensional andtransitiondynamicsunknown.

Individuals often make different decisions when faced with the same context, due to personal preferences and background. For instance, judges may vary in their leniency towards certain drug-related offenses, and doctors may vary in their preference for how to start treatment for certain types of patients.

Recently, such difficulty in multi-agent learning has been eased owing to the introduction ofcentralized training for decentralized execution(CTDE) [11, 45], which allows agents to access the global information andopponents' actions during thetraining phase.