Workshop Brings Together SoCal Leaders in Causal Reasoning

On February 6, 2025, around 80 researchers gathered at UC Irvine for the SoCal Workshop on Causal Reasoning. Through a series of research talks, speakers from universities across Southern California, including UCI, UCSD, UCLA, Caltech and USC, as well as Columbia University shared recent advancements in causal reasoning and discussed future directions. The workshop was organized as part of the second annual meeting of a $5 million, multi-institutional grant from the National Science Foundation on causal foundations of decision making and learning.

“The first annual grant meeting was held at Columbia University in New York, and this year, since it was held here at UCI, I thought it was a great opportunity to bring together causality experts here in Southern California,” says Computer Science Professor Rina Dechter. As one of five co-principal investigators on the NSF grant, she organized this year’s workshop at UCI’s Donald Bren School of Information and Computer Sciences (ICS).

Dechter was excited to have Judea Pearl from UCLA attend and speak. “He is a world-renowned leader in artificial intelligence and especially in causal inference, and he gave a philosophical talk on free will and AI,” she says. “There was a lot of interest in this talk.” Pearl also participated in a panel discussion moderated by Dechter.

Dechter was also pleased to have some of her ICS colleagues present their work. “There are two types of backgrounds for people who are doing causal reasoning, statistics and computer science, and we have a good representation from both.” (ICS is unique in that it houses both the Department of Computer Science and the Department of Statistics.) In addition, the workshop featured UCI faculty from the Department of Economics and the Department of Cognitive Science in the School of Social Sciences.

Dechter was particularly impressed with the talk by Professor Adrian Lozano-Duran of Caltech about the use of causal inference in aerospace research. “It was very interesting, coming from someone who is not necessarily adopting traditional approaches to causality, but rather developing methods that fit the domain within which he is working.”

This is exactly the sort of domain-based work the grant aims to support. “We’re facilitating collaborations with people from different backgrounds,” says Dechter.

The grant itself was motivated by use-cases work in robotics and public health. “Causality is very important in understanding the impact of patient treatment on recovery, especially in the context of sequential decision making, which is why we focus on advancing areas such as reinforcement learning and explainability,” says Dechter. “Another critical area is ethics and fairness. To address bias that may be present in AI systems, we need causal language that can articulate such notions in order to build more adequate AI tools.”

Dechter looks forward to future collaborations stemming from the workshop. “The event was good because it included talks beyond the typical areas of causality, hopefully inspiring people to find common ground with others in their research.”

If you missed the event, videos of the talks are available online.

— Shani Murray