Migration as a Probe: A Generalizable Benchmark Framework for Specialist vs. Generalist Machine-Learned Force Fields
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917021989470208 |
|---|---|
| author | Cao, Yi Clancy, Paulette |
| author_facet | Cao, Yi Clancy, Paulette |
| contents | Machine-learned force fields (MLFFs), especially pre-trained foundation models, are transforming computational materials science by enabling ab initio-level accuracy at molecular dynamics scales. Yet their rapid rise raises a key question: should researchers train specialist models from scratch, fine-tune generalist foundation models, or use hybrid approaches? The trade-offs in data efficiency, accuracy, cost, and robustness to out-of-distribution failure remain unclear. We introduce a benchmarking framework using defect migration pathways, evaluated through nudged elastic band trajectories, as diagnostic probes that test both interpolation and extrapolation. Using Cr-doped Sb2Te3 as a representative two-dimensional material, we benchmark multiple training paradigms within the MACE architecture across equilibrium, kinetic (atomic migration), and mechanical (interlayer sliding) tasks. Fine-tuned models substantially outperform from-scratch and zero-shot approaches for kinetic properties but show partial loss of long-range physics. Representational analysis reveals distinct, non-overlapping latent encodings, indicating that different training strategies learn different aspects of system physics. This framework provides practical guidelines for MLFF development and establishes migration-based probes as efficient diagnostics linking performance to learned representations, guiding future uncertainty-aware active learning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_00090 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Migration as a Probe: A Generalizable Benchmark Framework for Specialist vs. Generalist Machine-Learned Force Fields Cao, Yi Clancy, Paulette Chemical Physics Materials Science Machine Learning Computational Physics I.2.6; I.6.3; I.6.5; J.2 Machine-learned force fields (MLFFs), especially pre-trained foundation models, are transforming computational materials science by enabling ab initio-level accuracy at molecular dynamics scales. Yet their rapid rise raises a key question: should researchers train specialist models from scratch, fine-tune generalist foundation models, or use hybrid approaches? The trade-offs in data efficiency, accuracy, cost, and robustness to out-of-distribution failure remain unclear. We introduce a benchmarking framework using defect migration pathways, evaluated through nudged elastic band trajectories, as diagnostic probes that test both interpolation and extrapolation. Using Cr-doped Sb2Te3 as a representative two-dimensional material, we benchmark multiple training paradigms within the MACE architecture across equilibrium, kinetic (atomic migration), and mechanical (interlayer sliding) tasks. Fine-tuned models substantially outperform from-scratch and zero-shot approaches for kinetic properties but show partial loss of long-range physics. Representational analysis reveals distinct, non-overlapping latent encodings, indicating that different training strategies learn different aspects of system physics. This framework provides practical guidelines for MLFF development and establishes migration-based probes as efficient diagnostics linking performance to learned representations, guiding future uncertainty-aware active learning. |
| title | Migration as a Probe: A Generalizable Benchmark Framework for Specialist vs. Generalist Machine-Learned Force Fields |
| topic | Chemical Physics Materials Science Machine Learning Computational Physics I.2.6; I.6.3; I.6.5; J.2 |
| url | https://arxiv.org/abs/2509.00090 |