1 paper
Daniel S. Brown, Jordan Schneider, Anca D. Dragan +1
As humans interact with autonomous agents to perform increasingly complicated, potentially risky tasks, it is important to be able to efficiently evaluate an agent's performance an…