arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.06577physics.chem-ph

优化优化器:语言模型发现更快的分子弛豫方法

Optimizing the Optimizer: Language Models Discover Faster Molecular Relaxation Algorithms

Artem Tsypin, Vladimir Deshchenya, Kuzma Khrabrov, Denis Potapov, Maxim Radchenko, Artur Kadurin, Michael G. Medvedev

首次发表
浏览论文内容

中文总结 AI 辅助

本研究让语言模型通过自动研究重写几何优化器,生成AutoSella优化器家族,在保持能量降低的同时显著减少力调用次数,尤其在r2SCAN-3c水平上仅需Sella的40.2%–77.2%力调用。

中文摘要 AI 辅助

几何优化是许多量子化学工作流中的主要成本:每个优化步骤需要一次力评估,而在密度泛函理论水平上,该评估占主导地位。该领域的研究已产生了广泛的优化方法,我们探究语言模型能否通过自动研究改进其中最佳方法。一个智能体重写优化器本身以最小化力调用次数,并受两个准入门限制,这些门拒绝过早停止以及无法推广到未见分子的改进。从Sella(可用的最快开源优化器)出发,搜索产生了AutoSella,一个包含两个优化器的家族。两者在保留的分子基准和搜索中未使用的势能上,相对于Sella均实现了一致的力调用减少。最值得注意的是,在\ exttt{r2SCAN-3c}密度泛函理论水平上,最佳变体仅需Sella力调用次数的$40.2$--$77.2\%$即可实现相同的能量降低,尽管智能体未使用密度泛函理论梯度。

英文摘要

Geometry optimization is a major cost in many quantum-chemical workflows: each optimization step requires one force evaluation, and at the density-functional level that evaluation dominates the wall time. Research in this area has produced a broad range of optimization methods, and we ask whether a language model can improve on the best of them through autoresearch. An agent rewrites the optimizer itself to minimize force-call counts, restrained by two admission gates that reject premature stopping and improvements that do not generalize to unseen molecules. Starting from Sella, the fastest open-source optimizer available, the search produces AutoSella, a family of two optimizers. Both of them deliver consistent force-call reductions relative to Sella across held-out molecular benchmarks and potentials not used during the search. Most notably, at the r2SCAN-3c DFT level, the best variant requires only 40.2-77.2% of Sella's force calls while achieving the same energy reduction, even though agent used no DFT gradients.

发表机构

  • ITMO University(圣彼得堡国立信息技术机械与光学大学)
  • N. D. Zelinsky Institute of Organic Chemistry of Russian Academy of Sciences(俄罗斯科学院N·D·泽林斯基有机化学研究所)

机构由 AI 辅助整理,请以论文原文为准。

补充信息

↑