标题： 作为探针的迁移：掺杂材料中专业与通用机器学习力场的可推广基准框架 显示英文标题

标题： Migration as a Probe: A Generalizable Benchmark Framework for Specialist vs. Generalist Machine-Learned Force Fields in Doped Materials

摘要： 机器学习力场（MLFFs），特别是预训练的基础模型，有望将从头算级别的准确性带入分子动力学的长度和时间尺度。 然而，这种转变引发了一个核心问题：是应该从零开始构建专用模型，还是适应一个通用的基础模型以适用于特定系统？ 在数据效率、预测准确性和分布外（OOD）失败风险方面的权衡仍然不明确。 在这里，我们提出一个基准框架，在一个技术相关的二维材料测试案例中对比定制（从零开始）和微调的基础模型，即Cr插层的Sb2Te3，使用MACE架构。 我们的框架采用迁移路径，通过非弹性带（NEB）轨迹进行评估，作为诊断探针，测试插值和外推。 我们评估了平衡、动力学（原子迁移）和力学（层间滑动）任务的准确性。 虽然所有模型都能捕捉平衡结构，但对非平衡过程的预测出现分歧。 与从零开始和零样本模型相比，任务特定的微调显著提高了动力学准确性，但可能损害对长程物理的学习表示。 内部表示的分析表明，训练范式产生不同的、不重叠的系统物理潜在编码。 这项工作为MLFF开发提供了实用指南，强调基于迁移的探针作为高效的诊断工具，并提出了迈向不确定性感知主动学习策略的途径。 显示英文摘要

摘要： Machine-learned force fields (MLFFs), particularly pre-trained foundation models, promise to bring ab initio-level accuracy to the length and time scales of molecular dynamics. Yet this shift raises a central question: is it better to build a specialist model from scratch or adapt a generalist foundation model for a specific system? The trade-offs in data efficiency, predictive accuracy, and risks of out-of-distribution (OOD) failure remain unclear. Here, we present a benchmarking framework that contrasts bespoke (from scratch) and fine-tuned foundation models in a test case of a technologically relevant 2D material, Cr-intercalated Sb2Te3, using the MACE architecture. Our framework employs migration pathways, evaluated through nudged elastic band (NEB) trajectories, as a diagnostic probe that tests both interpolation and extrapolation. We assess accuracy for equilibrium, kinetic (atomic migration), and mechanical (interlayer sliding) tasks. While all models capture equilibrium structures, predictions for non-equilibrium processes diverge. Task-specific fine-tuning substantially improves kinetic accuracy compared with both from-scratch and zero-shot models, but can degrade learned representations of long-range physics. Analysis of internal representations shows that training paradigms yield distinct, non-overlapping latent encodings of system physics. This work offers a practical guide for MLFF development, highlights migration-based probes as efficient diagnostics, and suggests pathways toward uncertainty-aware active learning strategies.