Researchers · Interviewed at NeurIPS 2025
University of Chicago

NeurIPS 2025 · December 27, 2025 · 12 min
How do you scale causal discovery to tens of thousands of variables—without sacrificing theoretical guarantees?
In this #NeurIPS 2025 research highlight, Ashka Shah @UChicago presents a provably correct divide-and-conquer framework for causal discovery that tackles one of the field’s biggest bottlenecks: scalability.