HypergraphFormer: Learning Hypergraphs from LLMs for Editable Floor Plan Generation
Nikita KlimenkoAutodesk Research [email protected], Hesam SalehipourAutodesk Research [email protected], Parham EftekharYork University [email protected] Amir KhasahmadiAutodesk Research [email protected] Ramon Elias WeberUC Berkeley [email protected] · 2026
Read HypergraphFormer: Learning Hypergraphs from LLMs for Editable Floor Plan Generation by Nikita KlimenkoAutodesk Research [email protected], Hesam SalehipourAutodesk Research [email protected], Parham EftekharYork University [email protected] Amir KhasahmadiAutodesk Research [email protected] Ramon Elias WeberUC Berkeley [email protected] (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-2605-18932.txt.
Introduction
Nikita Klimenko and four colleagues generate floor plans with a language model that writes hypergraph descriptions of rooms and their connections. Trained by supervised fine-tuning, HypergraphFormer encodes spatial relationships and connectivity as text, then renders them as plans. On RPLAN and a new out-of-distribution set, it beats raster and vector methods and uses data more efficiently. Because footprints are decoupled from their subdivisions, the method handles irregular boundaries and stays highly editable for design work.
Catalog record
| CREATOR | Nikita KlimenkoAutodesk Research [email protected], Hesam SalehipourAutodesk Research [email protected], Parham EftekharYork University [email protected] Amir KhasahmadiAutodesk Research [email protected] Ramon Elias WeberUC Berkeley [email protected] |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 97243 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.18932
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
HypergraphFormer: Learning Hypergraphs from LLMs for Editable Floor Plan Generation Nikita KlimenkoAutodesk Research [email protected], Hesam SalehipourAutodesk Research [email protected], Parham EftekharYork University [email protected] Amir KhasahmadiAutodesk Research [email protected] Ramon Elias WeberUC Berkeley [email protected] Abstract In this work, we propose HypergraphFormer , a novel and efficient approach to floor plan generation based on learning hypergraph representations with a large language model (LLM). The model is trained via supervised fine-tuning to generate a hypergraph-based textual representation that encodes spatial relationships and connectivity information within floor plans. We train and evaluate our approach on the RPLAN dataset, and further demonstrate its generalizability on a separate out-of-distribution dataset, which we release in this paper. Our method outperforms state-of-the-art techniques based on rasterized or vectorized representations across a diverse set of metrics. We also show improved data efficiency, particularly under distribution shift. The hypergraph formulation enables the generation of floor plans for arbitrary, irregular, user-specified boundaries by decoupling apartment footprints from their functional and geometric subdivisions. Furthermore, we show that the proposed methodology offers a high degree of editability, making it particularly well suited to design-oriented workflows supported by LLMs. 1 1 footnotetext: Equal contribution. 2 2 footnotetext: Corresponding author. 3 3 footnotetext: This study was conducted during the author’s internship at Autodesk. 1 Introduction The design of architectural floor plans is challenging and time-consuming, requiring a manual, iterative process in which professional architects balance competing requirements [ 4 ] . Given the unprecedented scale of global urbanization, automated methods for floor plan analysis and generation are a key opportunity to make architectural design more accessible and to produce higher-quality indoor spaces at scale. Existing deep-learning approaches to floorplan generation commit to one of two output representations, both with structural drawbacks: raster-based methods generate floor plans as low-resolution pixel masks [ 13 , 14 , 22 , 34 , 3 , 7 ] , where the spatial resolution is bounded by the pixel size, walls and openings must be encoded implicitly, and the output requires non-trivial post-processing to recover a usable plan; vector-based methods predict per-room polygons or coordinate pairs [ 19 , 26 , 21 , 8 , 16 , 35 ] , but typically restrict the layout to axis-aligned rectangles inside a fixed boundary and still depend on geometric post-processing for closure and consistency. Beyond representation, a second axis matters in practice: the cost of editing a generated plan and the ability to generalize beyond the training distribution. As we summarize in Table 1 , every prior baseline either lacks fine-grained or boundary-level editability or recovers it only through dedicated training-time machinery and a full re-inference pass, none natively expresses arbitrary non-axis-aligned wall geometry, and none reports data-efficient training or out-of-distribution evaluation. Recent work has begun to leverage Large Language Models (LLMs) for layout synthesis [ 35 , 10 ] , but existing LLM-based pipelines hand the geometry off to a diffusion or raster decoder and inherit the same limitations. See Appendix A for a detailed discussion of related work. Table 1 : Feature comparison of representative floor plan generation methods. “Boundary” refers to the building outline (plus entrance/front-door when supplied). RC = room counts, RT = room types/labels, RS = room sizes/areas, RL = room locations/centers. For input/output columns, ✓ / ✗ denote whether the pathway is supported. For editability and extra-features columns, ✓ / ? / ✗ denote full, partial, and no support, respectively, as documented in the cited papers. Input Output Editability Extra Features Paper Access graph Adjacency graph Boundary Other Raster Vector Graph Boundary Editability Fine-grained Editability Irregular Boundaries Data Efficiency Out-of- Distribution House-GAN [ 13 ] ✗ ✓ ✗ – ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✗ House-GAN++ [ 14 ] ✓ ✗ ✗ – ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✗ Graph Transformer GANs [ 22 ] ✗ ✓ ✗ – ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✗ MaskPLAN [ 34 ] ✗ ✓ ✓ RT, RL, RS ✓ ✗ ✗ ? ? ✗ ✗ ✗ iPLAN [ 3 ] ✗ ✗ ✓ RT, RC, RL, RS ✓ ✗ ✗ ? ? ? ✗ ✗ WallPlan [ 21 ] ✗ ✓ ✓ – ✗ ✓ ✗ ? ✗ ✗ ✗ ✗ HouseDiffusion [ 19 ] ✓ ✗ ✗ – ✗ ✓ ✗ ✗ ✗ ? ✗ ✗ HouseTune [ 35 ] ✗ ✗ ✗ RT, RL, RS ✗ ✓ ✗ ✗ ✗ ✗ ✗ ✗ DiffPlanner [ 26 ] ✗ ✓ ✓ RC, RT, RS, RL ✗ ✓ ✗ ? ? ✗ ✗ ✗ GSDiff [ 8 ] ✗ ✓ ✓ – ✗ ✓ ✗ ? ✗ ? ✗ ✗ Graph2Plan [ 7 ] ✗ ✓ ✓ RC, RT, RL ✓ ✗ ✗ ? ? ✗ ✗ ✗ HypergraphFormer (Ours) ✓ ✗ ✗ – ✗ ✗ ✓ ✓ ✓ ✓ ✓ ✓ This paper takes an empirical stance: rather than proposing a new architecture or a new floor plan representation, we ask what becomes possible when a small LLM is trained to generate an existing structured representation. We adopt the graph-based textual representation of a floor plan introduced by 28 , referred to as a hypergraph , which decouples an apartment’s outer boundary from its interior layout by combining a binary space partition (BSP) tree, hierarchically representing the spatial decomposition into rooms, with an access graph capturing their functional connectivity. We show that a small instruction-tuned LLM can be fine-tuned to produce this representation directly, which unlocks a combination of properties that prior raster- and vector-based pipelines lack: native editability of the generated plan, strong out-of-distribution generalization to architect-designed apartments, and substantial data efficiency relative to state-of-the-art baselines. We refer to the resulting system as HypergraphFormer , and our contributions are: • Representation. We show that a lightweight LLM can be fine-tuned to generate the structured hypergraph representation of 28 directly, yielding a generative pipeline that produces floor plans conforming to arbitrary, user-specified boundaries without rasterization or per-room polygon prediction. • Editability. We show that this representation enables a family of fast, LLM-free procedural edits (e.g. adding or removing rooms, rotation/reflection, and gradient-descent refinement of room areas) that compose with the generated plan and further improve layout quality, as well as higher-level edits applied via LLM tool calls on the BSP tree and access graph. • Data efficiency. We conduct a controlled data-efficiency study showing that our approach matches the accuracy of state-of-the-art baselines using only a small fraction of their training data. • Out-of-distribution generalization. We demonstrate strong out-of-distribution generalization: trained on RPLAN, our model exceeds these baselines on WMR24 , our curated dataset of architect-designed floor plans whose distribution of apartment size, bedroom count, and boundary geometry differs substantially from the training distribution. We release WMR24 together with our training, inference, dataset-conversion, and procedural-editing code to enable full reproduction and extension of these results. 2 Methodology As introduced by 28 , a hypergraph is a reduced-order representation for floor plans: each apartment is decomposed into a boundary and a hypergraph stored as structured JSON, in which intermediate nodes encode binary space partitions of the interior and leaf nodes carry per-room semantics together with an access graph over door connections. We adopt this representation throughout and refer the reader to Appendix B for full details.