5. "LLM Routing with New Models Through Active Learning", presented by Alkiviadis Mertzios (MIT).

5/n, n = 5

3. "Too Much of a Good Thing? the Hidden Cost of Widespread AI-Generated Persuasion", presented by Yan Leng (University of Texas at Austin).
4. "Measuring Strategization in Recommendation: Users Adapt Their Behavior to Shape Future Content", presented by Hannah Li (Columbia University).

4/n

2. "Conversation Dynamics in Human-AI Collaborative Teams: Study of Group Decision-Making in Mixed-Agent Environments", presented by Jiehui Luo (University of Notre Dame).

3/n

1. "Agreeing to Disagree: Human-AI Collaboration in Ethical Decision-Making", presented by Caitlin Cunningham (University of Washington).

2/n

Congratulations!!

Congratulations!!

The video of the talk is now available online: youtu.be/gzPqQdo6vsU?...

GeomScale also participates in the Google Summer of Code program, which aims to help students contribute to large open-source projects. It is an amazing opportunity if you want to be involved with open-source scientific software!

FYI: github.com/GeomScale/gs...

Volesti is an open-source package for volume computation and sampling in high-dimensional constrained spaces.

High-dimensional constrained sampling has many applications in computational biology, finance, Bayesian inference, differential privacy, runtime verification, and many others.

Paper link: arxiv.org/pdf/2402.10659

Finally, we conduct a survey study involving human participants to assess our system's ability to aggregate preferences and reasoning. Our findings show that the system exhibits strong performance in both dimensions.

10/10

Our results indicate efficient coordination with reduced interactions between members and the LLM-based system. The system also effectively refines proposed options over time, ensuring their quality and equity.

9/n

We create synthetic employee profiles and simulate conversations at scale, leveraging LLMs to evaluate the system to avoid a costly real-user study.

8/n

Our system extracts individual preferences and suggests options that satisfy a significant portion of the members. We apply this system to corporate meeting scheduling.

7/n