Ofra
Amir

Agent Strategy Summarization: Describing Agent Behavior to People

Assistant professor at the faculty of industrial engineering and management, Technion

Ofra Amir

Ofra
Amir

Agent Strategy Summarization: Describing Agent Behavior to People

Assistant professor at the faculty of industrial engineering and management, Technion
Ofra Amir

Bio

Ofra Amir is an assistant professor at the faculty of industrial engineering and management at the Technion. She received her PhD in computer science from Harvard University, and holds a BSc and MSc in Information Systems Engineering, both from Ben-Gurion University. Before joining the Technion, Ofra was a post-doc and lecturer at the Harvard John A. Paulson School of Engineering and Applied Sciences. Ofra’s research combines Artificial Intelligence (AI) and Human-Computer Interaction (HCI), with the overarching goal of developing intelligent systems that support people in application domains of societal importance, such as healthcare and education. She is also interested in developing explainable AI methods that help people understand the behavior of AI agents, and in the ethics of AI systems. She is a recipient of J.P. Morgan Faculty Research Award (2019) and a Siebel Scholarship (2016), and was named one of Israel’s 40 under 40 by TheMarker magazine (2019).

Bio

Ofra Amir is an assistant professor at the faculty of industrial engineering and management at the Technion. She received her PhD in computer science from Harvard University, and holds a BSc and MSc in Information Systems Engineering, both from Ben-Gurion University. Before joining the Technion, Ofra was a post-doc and lecturer at the Harvard John A. Paulson School of Engineering and Applied Sciences. Ofra’s research combines Artificial Intelligence (AI) and Human-Computer Interaction (HCI), with the overarching goal of developing intelligent systems that support people in application domains of societal importance, such as healthcare and education. She is also interested in developing explainable AI methods that help people understand the behavior of AI agents, and in the ethics of AI systems. She is a recipient of J.P. Morgan Faculty Research Award (2019) and a Siebel Scholarship (2016), and was named one of Israel’s 40 under 40 by TheMarker magazine (2019).

Abstract

From cleaning robots to self-driving cars, autonomous and semi-autonomous agents are becoming increasingly prevalent. Understanding the capabilities and limitations of agents is important for users, as they might need to choose between different agents, adjust the level of autonomy of an agent, or work alongside an agent.

 

While prior work has developed methods for explaining individual decisions of an agent to a person retrospectively, these approaches do not provide users with a global understanding of an agent’s expected behavior in a range of situations. In this talk, I will discuss our recent work on agent strategy summarization, which aims to convey to people the policy of an agent by demonstrating its behavior in a carefully selected set of world-states.

 

I will show empirical results demonstrating that summaries can help people assess agents’ capabilities, but also that there are many subtleties to the way in which people interpret summaries.

Abstract

From cleaning robots to self-driving cars, autonomous and semi-autonomous agents are becoming increasingly prevalent. Understanding the capabilities and limitations of agents is important for users, as they might need to choose between different agents, adjust the level of autonomy of an agent, or work alongside an agent.

 

While prior work has developed methods for explaining individual decisions of an agent to a person retrospectively, these approaches do not provide users with a global understanding of an agent’s expected behavior in a range of situations. In this talk, I will discuss our recent work on agent strategy summarization, which aims to convey to people the policy of an agent by demonstrating its behavior in a carefully selected set of world-states.

 

I will show empirical results demonstrating that summaries can help people assess agents’ capabilities, but also that there are many subtleties to the way in which people interpret summaries.

Planned Agenda

Planned Agenda