Robin Na (@RobinNa_) profile picture

PhD candidate @MIT | AI for social science, multi-agent AI | formerly @AmazonScience, @CarnegieMellon

Cambridge, MA
Joined March 2019
Photo by Robin Na (@RobinNa_)Photo by Robin Na (@RobinNa_)Photo by Robin Na (@RobinNa_)Photo by Robin Na (@RobinNa_)Photo by Robin Na (@RobinNa_)
Prompting LLMs to behave like people can improve simulations, but how should we design this? What info is needed, and how much? With @Kehang_Zhu @locross @bakkermichiel, we build “behavioral profiles” that help LLMs simulate human behavior under distribution shifts. 1/8 🧵
7
24
2
144
13,751
2/8 Human behavior is complex, so any person can be described by countless features. We can’t fit them all into a prompt. Even if we could, including every detail might not improve the simulation. It’s also hard to predict which features will matter most.
1
5
537