JanusPipe introduces SymFold and WaveK to enable efficient 3D-parallel training for conservative MLIPs, reporting 1.51x and 1.45x average throughput gains over 1F1B and Hanayo baselines on 32 GPUs.
arXiv preprint arXiv:2504.16068 , year=
8 Pith papers cite this work, alongside 11 external citations. Polarity classification is still indexing.
8
Pith papers citing it
11
external citations · external index
citation-role summary
background 3
citation-polarity summary
verdicts
UNVERDICTED 8roles
background 3representative citing papers
Physics-constrained equivariant GNN warm starts cut DFT+DMFT self-consistency iterations by 2-4x across Fe, FeO and NiO, enabling an MLIP-based NVE coexistence simulation that yields a 6225 K melting temperature for hcp-Fe at 330 GPa.
Sparsity-promoting fine-tuning adapts equivariant materials foundation models by selectively updating ~3% of parameters to match full fine-tuning on molecular and crystalline benchmarks while revealing interpretable physical patterns.
An equivariant GNN trained on DFT energies of solvent-free PGNs reproduces those energies at low cost and drives MC simulations that match experimental structures despite training exclusively on out-of-equilibrium data.
Subsurface vacancies are thermodynamically preferred over surface vacancies on specific FCC and HCP metal surfaces due to a geometry-electronic decoupling mechanism identified via DFT and ML force fields.
Structural pruning of SO(3) equivariant atomistic models from large checkpoints yields 1.5-4x fewer parameters and 2.5-4x less pre-training compute than small models trained from scratch, while outperforming them on most Matbench Discovery metrics and downstream tasks.
This perspective article develops a definition of foundational MLIPs and poses six open questions that the authors believe will define future research in machine-learned interatomic potentials.
Machine learning models trained on quantum mechanical data can predict defect properties in solids with high accuracy but at much lower computational cost than traditional methods.
citing papers explorer
-
JanusPipe: Efficient Pipeline Parallel Training for Machine Learning Interatomic Potentials
JanusPipe introduces SymFold and WaveK to enable efficient 3D-parallel training for conservative MLIPs, reporting 1.51x and 1.45x average throughput gains over 1F1B and Hanayo baselines on 32 GPUs.
-
Physics-Constrained Self-Energy Warm Starts for Charge-Self-Consistent DFT+DMFT: Application to Iron at Core Conditions
Physics-constrained equivariant GNN warm starts cut DFT+DMFT self-consistency iterations by 2-4x across Fe, FeO and NiO, enabling an MLIP-based NVE coexistence simulation that yields a 6225 K melting temperature for hcp-Fe at 330 GPa.
-
Robust and Interpretable Adaptation of Equivariant Materials Foundation Models via Sparsity-promoting Fine-tuning
Sparsity-promoting fine-tuning adapts equivariant materials foundation models by selectively updating ~3% of parameters to match full fine-tuning on molecular and crystalline benchmarks while revealing interpretable physical patterns.
-
Using graph neural networks to predict many-body interactions in amorphous materials
An equivariant GNN trained on DFT energies of solvent-free PGNs reproduces those energies at low cost and drives MC simulations that match experimental structures despite training exclusively on out-of-equilibrium data.
-
Anomalous Subsurface Vacancy Stabilization Dictated by Geometry-Electronic Decoupling on Metal Surfaces
Subsurface vacancies are thermodynamically preferred over surface vacancies on specific FCC and HCP metal surfaces due to a geometry-electronic decoupling mechanism identified via DFT and ML force fields.
-
Compact SO(3) Equivariant Atomistic Foundation Models via Structural Pruning
Structural pruning of SO(3) equivariant atomistic models from large checkpoints yields 1.5-4x fewer parameters and 2.5-4x less pre-training compute than small models trained from scratch, while outperforming them on most Matbench Discovery metrics and downstream tasks.
-
Six Open Questions in Machine-Learned Interatomic Potential Foundation Models
This perspective article develops a definition of foundational MLIPs and poses six open questions that the authors believe will define future research in machine-learned interatomic potentials.
-
Accelerating point defect simulations using data-driven and machine learning approaches
Machine learning models trained on quantum mechanical data can predict defect properties in solids with high accuracy but at much lower computational cost than traditional methods.