Hilbert-Geo: Solving Solid Geometric Problems by Neural-Symbolic Reasoning
Ruoran Xu, Haoyu Cheng, Bin Dong, Qiufeng Wang · 2026
Read Hilbert-Geo: Solving Solid Geometric Problems by Neural-Symbolic Reasoning by Ruoran Xu, Haoyu Cheng, Bin Dong, Qiufeng Wang (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-2605-16385.txt.
Description
Xu and three colleagues show their framework travels: it reaches state-of-the-art accuracy of 80.2 percent on PlaneFGeo3k, a plane-geometry set. The claim is Hilbert-Geo, a unified formal-language framework for solid geometry, the first of its kind. Its Parse2Reason route turns problem text and diagrams into conditional description language, built from predicates stating geometric conditions. It then uses that language and the theorem bank for relational inference and algebraic computation. On SolidFGeo2k it hits 77.3 percent and on MathVerse-Solid 84.1 percent.
Catalog record
| CREATOR | Ruoran Xu, Haoyu Cheng, Bin Dong, Qiufeng Wang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 71786 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.16385