TY - GEN
T1 - A Unifying Bayesian Formulation of Measures of Interpretability in Human-AI Interaction
AU - Sreedharan, Sarath
AU - Kulkarni, Anagha
AU - Smith, David E.
AU - Kambhampati, Subbarao
N1 - Funding Information:
We would like to thank Dr. Tathagata Chakraborti for his significant contributions to an earlier version of this paper and for extensive discussions on the topic. This research is supported in part by ONR grants N00014-16-1-2892, N00014-18-1-2442, N00014-18-1-2840, N00014-9-1-2119, AFOSR grant FA9550-18-1-0067, DARPA SAIL-ON grant W911NF-19-2-0006, NASA grant NNX17AD06G, and J.P. Morgan Faculty Research Award.
Publisher Copyright:
© 2021 International Joint Conferences on Artificial Intelligence. All rights reserved.
PY - 2021
Y1 - 2021
N2 - Existing approaches for generating human-aware agent behaviors have considered different measures of interpretability in isolation. Further, these measures have been studied under differing assumptions, thus precluding the possibility of designing a single framework that captures these measures under the same assumptions. In this paper, we present a unifying Bayesian framework that models a human observer's evolving beliefs about an agent and thereby define the problem of Generalized Human-Aware Planning. We will show that the definitions of interpretability measures like explicability, legibility and predictability from the prior literature fall out as special cases of our general framework. Through this framework, we also bring a previously ignored fact to light that the human-robot interactions are in effect open-world problems, with respect to the human's beliefs about the agent. The human may hold beliefs unknown to the agent and may also form new hypotheses about the agent when presented with novel or unexpected behaviors.
AB - Existing approaches for generating human-aware agent behaviors have considered different measures of interpretability in isolation. Further, these measures have been studied under differing assumptions, thus precluding the possibility of designing a single framework that captures these measures under the same assumptions. In this paper, we present a unifying Bayesian framework that models a human observer's evolving beliefs about an agent and thereby define the problem of Generalized Human-Aware Planning. We will show that the definitions of interpretability measures like explicability, legibility and predictability from the prior literature fall out as special cases of our general framework. Through this framework, we also bring a previously ignored fact to light that the human-robot interactions are in effect open-world problems, with respect to the human's beliefs about the agent. The human may hold beliefs unknown to the agent and may also form new hypotheses about the agent when presented with novel or unexpected behaviors.
UR - http://www.scopus.com/inward/record.url?scp=85115827693&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85115827693&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85115827693
T3 - IJCAI International Joint Conference on Artificial Intelligence
SP - 4602
EP - 4610
BT - Proceedings of the 30th International Joint Conference on Artificial Intelligence, IJCAI 2021
A2 - Zhou, Zhi-Hua
PB - International Joint Conferences on Artificial Intelligence
T2 - 30th International Joint Conference on Artificial Intelligence, IJCAI 2021
Y2 - 19 August 2021 through 27 August 2021
ER -