aiqbench analysis
This research demonstrates a generative AI system that solves complex, unsatisfiable material design problems by diagnosing constraint conflicts rather than merely returning the nearest imperfect match. By generating inclusion-minimal unsatisfiable subsets and specific repair slacks, the model ensures engineering feasibility where traditional nearest-neighbor approaches fail to respect critical cost or density limits. The work establishes a rigorous property-registry contract that prevents hallucinated parameters from reaching the search engine, significantly advancing reliable AI-assisted materials discovery.
A property-registry contract for retrieve-or-refuse thermal-mechanical lattice search
arXiv:2609.14741v1 Announce Type: cross Abstract: Early thermal-mechanical lattice requirements are knowledge-intensive and often jointly unsatisfiable: an engineer asks for a cell that is light, stiff, laterally conducting and cheap, and no cell in the library satisfies it. A design system should say so, and say which requirement to loosen and by how much, rather than return the nearest row. A generative model can return a candidate even when the library holds none. This work applies established conflict diagnosis to a catalogue of homogenised properties. Search over 1,397 homogenised unit cells and 19 base materials (26,543 combinations) returns a catalogue row that a second solver can rebuild, or, for an empty feasible set of at most eight constraints, inclusion-minimal unsatisfiable subsets (MUS) and the slack of a repair, not a neighbour. One registry declaration generates the prompt and the evaluator, so undeclared parsed keys cannot reach search. On a frozen suite of 64 typed queries, retrieve-or-refuse matches min-repair on every feasible query (48/48) and refuses every empty one (16/16) with MUS and slack. Constraint-ignoring nearest-neighbour and penalty search violate stated cost or density limits. On 216 empty queries, repairs printed at three significant figures and rounded outward stay feasible in every case (216/216). When density and cost must be kept, a list of minimum-cardinality repairs keeps them on 209 of 211 queries; a protection-first repair keeps them on all 211, the same as full diagnosis. A 308-request parse benchmark is a template-text check, not unconstrained engineer prose.
Abstract:Early thermal-mechanical lattice requirements are knowledge-intensive and often jointly unsatisfiable: an engineer asks for a cell that is light, stiff, laterally conducting and cheap, and no cell in the library satisfies it. A design system should say so, and say which requirement to loosen and by how much, rather than return the nearest row. A generative model can return a candidate even when the library holds none. This work applies established conflict diagnosis to a catalogue of homogenised properties. Search over 1,397 homogenised unit cells and 19 base materials (26,543 combinations) returns a catalogue row that a second solver can rebuild, or, for an empty feasible set of at most eight constraints, inclusion-minimal unsatisfiable subsets (MUS) and the slack of a repair, not a neighbour. One registry declaration generates the prompt and the evaluator, so undeclared parsed keys cannot reach search. On a frozen suite of 64 typed queries, retrieve-or-refuse matches min-repair on every feasible query (48/48) and refuses every empty one (16/16) with MUS and slack. Constraint-ignoring nearest-neighbour and penalty search violate stated cost or density limits. On 216 empty queries, repairs printed at three significant figures and rounded outward stay feasible in every case (216/216). When density and cost must be kept, a list of minimum-cardinality repairs keeps them on 209 of 211 queries; a protection-first repair keeps them on all 211, the same as full diagnosis. A 308-request parse benchmark is a template-text check, not unconstrained engineer prose.
30 pages, 7 figures, 22 tables. Appendices A-D give the full property registry, the generated language-model prompt, the frozen 64-query suite, and one worked tool transcript
Computational Engineering, Finance, and Science (cs.CE); Materials Science (cond-mat.mtrl-sci); Artificial Intelligence (cs.AI)
arXiv:2609.14741 [cs.CE]
(or arXiv:2609.14741v1 [cs.CE] for this version)
https://doi.org/10.48550/arXiv.2609.14741 arXiv-issued DOI via DataCite (pending registration)
Full text under the publisher's open licence · 294 words
Related coverage
Article: The Platform Engineering Playbook for Production LLMs
InfoQ · Oct 5, 2026
News
Quantum Computing Report · Jun 7, 2026
LiveLaw Supreme Court Half-Yearly Digest: January - June 2026 [Part 2]
livelaw.in · Sep 13, 2026
Lattice-Based Cryptography Explained
Quantum Zeitgeist · Aug 15, 2026