ArrowFlow: Hierarchical Machine Learning in the Space of Permutations
Ozgur Yilmaz · 2026
Read ArrowFlow: Hierarchical Machine Learning in the Space of Permutations by Ozgur Yilmaz (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-2604-04087.txt.
Description
Ozgur Yilmaz introduced ArrowFlow, a machine learning architecture working entirely in permutation space. Its units are ranking filters: learned orderings comparing inputs by Spearman's footrule distance and updating by rank-matrix accumulation, a non-gradient rule. Layers stack hierarchically, with each layer's output ranking feeding the next, enabling deep ordinal learning with no floating-point parameters at the core. ArrowFlow beat all baselines on Iris and stayed competitive on most UCI sets, linking fairness-axiom violations to useful inductive biases.
Catalog record
| CREATOR | Ozgur Yilmaz |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 108421 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.04087