Spyra-20B: A Proof-of-Concept for Explicit Architectural Design Reasoning in Domain-Specialised LLMs
Smart Design Policies · 15.08.2026 · 3 (1) · 14-28
DOI: 10.38027/smart.v3n1-2 ↗Kurz gefasst
Fachartikel in der Zeitschrift Smart Design Policies, online erschienen am 15.08.2026. Verfasst von Nik Ansre und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Domain-specifically fine-tuned language models (BloombergGPT, Med-PaLM) have proven a viable route to adapting vocabulary and specialist knowledge to particular fields of application. For architectural practice, however, the central challenge is not terminology but the deliberative structure of the design decision: the weighing of competing objectives across building law, quality of use, cost and sustainability. This paper presents Spyra-20B, a proof-of-concept produced by QLoRA fine-tuning of gpt-oss-20B on 1,693 architecture- and planning-specific dialogue examples. Two features characterise the approach: a three-level deliberation depth controllable by metadata (reasoning_effort), and a strict separation of analysis and answer channels that renders the deliberative process inspectable. An evaluation on 50 law- and planning-related items across four conditions (base model vs. Spyra-20B, closed-book vs. open-book with RAG; 4,000 requests) yields a negative headline result: the fine-tuned model attains roughly six percentage points lower answer accuracy (Spyra 61.1% / 70.4%, base 67.3% / 76.3%; p < 0.001), while retrieval improves both models by about nine percentage points. Fine-tuning changes the form of the output: more schematic, twice as fast at the median, but with a sixfold increase in unrecoverable format failures. We derive four governance requirements: reasoning transparency, external factual grounding, local executability and multi-sample verification, and conclude that such systems should be assessed along architectural properties rather than scores on individual benchmarks.
Zitieren
APA
Ansre, N., Hiresekorn, Y., & Grunwald, G. (2026). Spyra-20B: A Proof-of-Concept for Explicit Architectural Design Reasoning in Domain-Specialised LLMs. Smart Design Policies, 3(1), 14-28. https://doi.org/10.38027/smart.v3n1-2
Vancouver
Ansre N, Hiresekorn Y, Grunwald G. Spyra-20B: A Proof-of-Concept for Explicit Architectural Design Reasoning in Domain-Specialised LLMs. Smart Design Policies. 2026;3(1):14-28. doi:10.38027/smart.v3n1-2
BibTeX
@article{Ansre2026,
author = {Ansre, Nik and Hiresekorn, Yara and Grunwald, Gregor},
title = {Spyra-20B: A Proof-of-Concept for Explicit Architectural Design Reasoning in Domain-Specialised LLMs},
journal = {Smart Design Policies},
year = {2026},
volume = {3},
number = {1},
pages = {14--28},
doi = {10.38027/smart.v3n1-2}
}
EUVIMED