arXiv：cs.AI· Dongyeop Woo, Nofit Segal, Nayoung Kim, Minkyu Kim, Rafael Gomez-Bombarelli, Sungsoo Ahn·· 16 小时前
PXRD-Bench：从粉末衍射确定晶体结构还有多远？
PXRD-Bench: How Close Are We to Determining Crystal Structures from Powder Diffraction?
AI 导读
研究者提出统一基准 PXRD-Bench，将粉末 X 射线衍射（PXRD）到晶体结构的评估扩展到未知元素原子数与双相混合物结构回收，提供约 410 万个带模拟 PXRD 的训练结构，并测试对未见组分与结构原型的泛化。
正文
Abstract:Determining crystal structures from powder X-ray diffraction (PXRD) would enable routine characterization of newly synthesized materials, including those absent from reference databases. Generative models offer a promising approach, but existing evaluations remain fragmented and largely assume single-phase samples with known composition. To fill this gap, we introduce PXRD-Bench, a unified benchmark that extends evaluation to unknown per-element atom counts and recovery of both structures in two-phase mixtures. It provides approximately 4.1 million training structures with simulated PXRD, tests generalization to unseen compositions and structural prototypes, and evaluates recovery from experimental PXRD after excluding structurally matching training entries. To provide a strong reference baseline, we also introduce Bragglite, a PXRD-to-structure model developed through a systematic study of PXRD modeling choices. It achieves a 44.2\% Top-1 match rate on experimental RRUFF patterns with known composition, 24.6\% above the strongest prior baseline. Further analysis yields several key findings: some models fail to benefit from experimental PXRD, while generalization to unseen structural prototypes and recovery from two-phase mixtures remain challenging. Together, our benchmark, systematic studies, and Bragglite clarify the capabilities and limitations of current PXRD-to-structure models.
| Subjects: | Artificial Intelligence (cs.AI); Materials Science (cond-mat.mtrl-sci) |
| Cite as: | arXiv:2610.16480 [cs.AI] |
| (or arXiv:2610.16480v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.16480 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Dongyeop Woo [view email]
[v1]
Mon, 12 Oct 2026 09:34:51 UTC (1,747 KB)
来源：arXiv：cs.AI · arxiv.org