Explicability? Legibility? Predictability? Transparency? Privacy? Security? The Emerging Landscape of Interpretable Agent Behavior
This work addresses a foundational problem for researchers in AI and robotics by organizing an emerging field, though it is incremental as it synthesizes existing ideas rather than introducing new methods.
The paper tackles the lack of coherence in the field of interpretable agent behavior by providing a taxonomy to clarify overlapping and conflicting concepts such as explicability, legibility, predictability, transparency, privacy, and security.
There has been significant interest of late in generating behavior of agents that is interpretable to the human (observer) in the loop. However, the work in this area has typically lacked coherence on the topic, with proposed solutions for "explicable", "legible", "predictable" and "transparent" planning with overlapping, and sometimes conflicting, semantics all aimed at some notion of understanding what intentions the observer will ascribe to an agent by observing its behavior. This is also true for the recent works on "security" and "privacy" of plans which are also trying to answer the same question, but from the opposite point of view -- i.e. when the agent is trying to hide instead of revealing its intentions. This paper attempts to provide a workable taxonomy of relevant concepts in this exciting and emerging field of inquiry.