From Rashomon Theory to PRAXIS: Efficient Decision Tree Rashomon Sets
Zakk Heile, Hayden McTavish, Varun Babbar, Margo Seltzer, Cynthia Rudin · 2026
Read From Rashomon Theory to PRAXIS: Efficient Decision Tree Rashomon Sets by Zakk Heile, Hayden McTavish, Varun Babbar, Margo Seltzer, Cynthia Rudin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-00202.txt.
Introduction
Heile and co-authors speed the hunt for sets of near-best choice trees. Many models can tie for best on the same data. That whole set is called a Rashomon set. Users gain from seeing the full set, since taste and field lore can pick a winner. Yet listing the set eats vast time and memory.
Their tool, PRAXIS, takes a close sketch of the full set. It cuts run time and memory by huge steps. Tests show it finds near all of the true set in most runs. The tool works on real data sets at field scale. Code is free for all to use. The work opens set-based choice to teams that could not afford the old cost.
Catalog record
| CREATOR | Zakk Heile, Hayden McTavish, Varun Babbar, Margo Seltzer, Cynthia Rudin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 280 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.00202