Nearly one in five near-hull Crystoria structures showed harmonic instability
Source check: Source checked. Not independent scientific confirmation.
Finding: An EquFlashV2 analysis found that 187,983 of 980,560 structures within 10 meV per atom of the model’s reference hull still had unstable harmonic modes. This is a computational result for the curated Crystoria collection, rather than a direct measurement of crystal stability.
Technology: EquFlashV2 interatomic potential; harmonic phonon calculations
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- EquFlashV2 predicts forces for structural relaxation and displaced-structure harmonic phonon calculations, and supplies structure energies for positioning on a reference hull.
- Test and result
- In the curated Crystoria population, 187,983 of 980,560 structures within 10 meV/atom of the reference hull remained dynamically unstable under the authors' harmonic criterion. Independent fixed-geometry DFT phonon comparison used 9,958 nonmagnetic semiconductors.
- Limitation / still unknown
- The comparison covers an energy-selected collection that includes generated inversion partners sharing scalar calculations with parents. Harmonic modes do not account for finite-temperature stabilization. The external DFT comparison used fixed structures and only nonmagnetic semiconductors.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://arxiv.org/abs/2610.38195v1 · https://arxiv.org/html/2610.38195v1 · https://zenodo.org/records/23572554
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 21 Nov 2026. Review scope: arXiv v1 §§2, 4.2 and 5, Supplement S2, and dated Zenodo 1.0.0 source metadata.
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.