A walkthrough of the Potential Outcomes Framework from first principles, exploring why causal claims are harder than they look and how social scientists build credible arguments.
Five Penn physicists — from HEP theorists to experimentalists to astrophysicists — share their real AI workflows, covering symbolic computation, agentic coding, and the honest limits of trust.
Postdoctoral researchers across two sessions compared model selection strategies, built MCP servers live, and debated what agentic AI means for peer review integrity and student learning.
A narrative essay on how to keep sequence and evidence visible under polarization: turning fast-moving crises into structured, revisitable context.
A structured workflow for completing metadata without surrendering control.
Using machine learning to uncover where the most predictive insights hide in data.