发表机构
University of North Carolina at Charlotte; North Carolina Battery Complexity, Autonomous Vehicle and Electrification (BATT CAVE) Research Center（北卡罗来纳大学夏洛特分校; 北卡罗来纳州电池复杂性、自动驾驶与电气化（BATT CAVE）研究中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出CrystalGRPO框架，通过强化学习后训练优化基于流模型的晶体结构预测，两种模式在多数据集骨干组合中提升了恢复性能与覆盖率。
AI 中文摘要
基于流的生成模型可高效生成晶体结构预测（CSP）的候选结构，但其预训练目标未直接优化下游目标的恢复。强化学习后训练提供了灵活解决方案，但现有方法主要依赖能量奖励和仅坐标的随机策略。预测能量无法识别参考多型体，而奖励驱动的集中会降低Top-N恢复所需的候选覆盖率。我们提出CrystalGRPO，一种CSP对齐的后训练框架，将现有ODE-to-SDE策略构造扩展到联合坐标-晶格状态。CrystalGRPO结合MACE预测的能量与基于StructureMatcher的恢复评分，提供两种运行模式：CrystalGRPO-Q优先单样本恢复，CrystalGRPO-C结合全轨迹参考正则化与感知覆盖率的组优势，以保持有限预算下的目标恢复。在MP-20和MPTS-52数据集上，采用PXRDGen和OMatG骨干网络，两种变体在所有四个骨干-数据集设置中均降低了仅坐标强化学习的1样本和20样本RMSE；CrystalGRPO-Q持续提升Top-1，CrystalGRPO-C在所有设置中实现更高的Top-20。
英文摘要
Flow-based generative models can efficiently produce candidate structures for crystal structure prediction (CSP), but their pretrained objectives do not directly optimize downstream target recovery. Reinforcement-learning post-training offers a flexible solution, yet existing approaches rely primarily on energy rewards and coordinate-only stochastic policies. Predicted energy does not identify the reference polymorph, while reward-driven concentration can reduce the candidate coverage required for Top-N recovery. We introduce CrystalGRPO, a CSP-aligned post-training framework that extends existing ODE-to-SDE policy constructions to the joint coordinate--lattice state. CrystalGRPO combines MACE-predicted energy with a StructureMatcher-based recovery score and provides two operating modes: CrystalGRPO-Q, which prioritizes single-draw recovery, and CrystalGRPO-C, which combines full-trajectory reference regularization with a coverage-aware group advantage to preserve finite-budget target recovery. Across MP-20 and MPTS-52 with PXRDGen and OMatG backbones, both variants reduce one- and twenty-sample RMSE relative to coordinate-only reinforcement in all four backbone--dataset settings. CrystalGRPO-Q consistently improves Top-1, whereas CrystalGRPO-C achieves a higher Top-20 across all settings.
CommentsAdd acknowledgments
