Reproduce “Graph Distance Based on Cause-Effect Estimands with Latents” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zhufeng Li, Niki Kilbertus
The authors introduce the Fixing Identification Distance (FID), which compares graphs by intervention formulas they imply, including graphs with latent confounding. It canonicalizes fixing-based formulas and scores their overlap across treatment–outcome pairs. Synthetic comparisons show FID varies among graphs with the same edge-edit score and correlates with separation distance (r = 0.292); it is not an evaluation of estimates.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.