Pith. sign in

REVIEW 6 cited by

Fine-tuning foundation models of materials interatomic potentials with frozen transfer learning

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2502.15582 v3 pith:H5ZKP2FS submitted 2025-02-21 cond-mat.mtrl-sci

Fine-tuning foundation models of materials interatomic potentials with frozen transfer learning

classification cond-mat.mtrl-sci
keywords potentialsdatafoundationlearningmodelstransferaccuracyaccurate
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Machine-learned interatomic potentials are revolutionising atomistic materials simulations by providing accurate and scalable predictions within the scope covered by the training data. However, generation of an accurate and robust training data set remains a challenge, often requiring thousands of first-principles calculations to achieve high accuracy. Foundation models have started to emerge with the ambition to create universally applicable potentials across a wide range of materials. While foundation models can be robust and transferable, they do not yet achieve the accuracy required to predict reaction barriers, phase transitions, and material stability. This work demonstrates that foundation model potentials can reach chemical accuracy when fine-tuned using transfer learning with partially frozen weights and biases. For two challenging datasets on reactive chemistry at surfaces and stability and elastic properties of tertiary alloys, we show that frozen transfer learning with 10-20% of the data (hundreds of datapoints) achieves similar accuracies to models trained from scratch (on thousands of datapoints). Moreover, we show that an equally accurate, but significantly more efficient surrogate model can be built using the transfer learned potential as the ground truth. In combination, we present a simulation workflow for machine learning potentials that improves data efficiency and computational efficiency.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Dyna-Mat: End-to-end benchmarking of foundation machine learning interatomic potentials in finite-temperature ensembles

    cond-mat.mtrl-sci 2026-07 conditional novelty 6.5

    Dyna-Mat-v1.0 benchmarks 15 foundation MLIPs on finite-T MD observables, finding average force-error correlation with RDF/VDOS but systematic pressure failures and near-Pareto optimality of latest cross-trained models.

  2. Distilling first-principles accuracy into compact machine learning potentials for condensed-phase chemistry

    physics.chem-ph 2026-06 unverdicted novelty 6.0

    Distilled compact MLIPs from transfer-learned teachers reproduce observables more reliably than same-size models trained directly and enable practical PIMD umbrella sampling of water dissociation at TiO2 interface wit...

  3. VibroML: an automated toolkit for high-throughput vibrational analysis and dynamic instability remediation of crystalline materials using machine-learned potentials

    cond-mat.mtrl-sci 2026-04 unverdicted novelty 6.0

    VibroML automates remediation of dynamic instabilities in crystalline materials by combining MLIPs with genetic algorithms for polymorph search, finite-temperature MD validation, and compositional alloying to yield st...

  4. Revisiting quantum effects on dislocation glide in bcc metals from DFT calculations and machine-learning potentials

    cond-mat.mtrl-sci 2026-06 unverdicted novelty 5.0

    Quantum zero-point effects contribute less to low-temperature dislocation glide in bcc metals than earlier empirical-potential studies indicated, leaving the simulation-experiment discrepancy unresolved.

  5. Fine-tuning MLIP foundation models: strategies for accuracy and transferability

    physics.chem-ph 2026-06 unverdicted novelty 5.0

    Systematic tests show naive fine-tuning excels for single-task accuracy while multihead replay best preserves out-of-distribution robustness in MLIP adaptation.

  6. Machine-Learned Force Fields for Lattice Dynamics at Coupled-Cluster Level Accuracy

    cond-mat.mtrl-sci 2025-07 unverdicted novelty 5.0

    Machine-learned force fields trained on coupled-cluster potential energy surfaces produce phonon dispersions and vibrational densities of states for solids that agree better with experiment than DFT-based models.