Learning to Incentivize Other Learning Agents

Much of this effort has focused on the single-agent setting, in which an agent maximizes a predefined extrinsic reward function.