Papers › From Rashomon Theory to PRAXIS: Efficient Decision Tree Rashomon Sets

From Rashomon Theory to PRAXIS: Efficient Decision Tree Rashomon Sets

29 May 2026ICMLarXiv:2606.00202added by Syntology

Zakk Heile, Hayden McTavish, Varun Babbar, Margo Seltzer, Cynthia Rudin

Title, abstract, authors and date from arXiv's metadata (CC0); this paper is not in the Papers with Code archive (frozen 2025-07-28).

Standard machine learning pipelines often admit many near-optimal models. These "Rashomon sets" pose a range of challenges and opportunities for uncertainty-aware, robust decision making. They allow users to incorporate domain knowledge and preferences that would otherwise be difficult to specify directly in an objective, and they quantify diversity among valid models for a given training dataset and objective function. However, computation of Rashomon sets, even for simple, interpretable model classes such as sparse decision trees, continues to require immense memory and runtime resources. We present PRAXIS, an algorithm to approximate this Rashomon set with orders of magnitude improvement in runtime and memory usage. We validate that PRAXIS regularly recovers almost all of the full Rashomon set. PRAXIS allows researchers and practitioners to scalably model the Rashomon set for real-world datasets. Code for PRAXIS is available at https://github.com/zakk-h/PRAXIS

Code

Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Syntology holds the repository link but has not harvested or run code from it.

Results from the paper

The Papers with Code archive ends with its 2025-07-28 snapshot. This paper's arXiv identifier, 2606.00202, was issued in June 2026, after that date, so the archive has no leaderboard rows for it.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections