Skip to content

Screen candidates with MACE

pip install "meidnet[stability]"            # ase + mace-torch (~500 MB of weights on first use)
meidnet screen runs/my_oxides/generation --train-csv materials.csv

For each CIF: relax cell and positions with the MACE-MP-0 universal potential, compute the formation energy per atom against elemental reference phases, and report

  • S passes the formation-energy filter — ΔH_f ≤ 0.10 eV/atom against the elemental phases (--threshold); the report prints it as STABLE,
  • U unique — no structural duplicate among the candidates,
  • N novel — reduced formula absent from the training table.

Results go to stability.csv in the folder. Structures with |ΔH_f| > 5 eV/atom are flagged as artefacts and excluded from the rates. The reference phases and GGA-style O/N corrections are those of the paper.

The S is not stability in the usual sense. The common 0.1 eV/atom threshold for metastable, possibly synthesisable materials concerns the energy above the convex hull of all competing phases; the formation energy against the elements is a weaker test, so a candidate can pass it and still decompose into other compounds. A stability statement needs the energy above the hull (DFT, or an MLIP against a reference set of competing phases).

This is a first filter, not a verdict: MACE formation energies carry ~0.1 eV/atom uncertainty, and a cubic prototype may relax to a lower-symmetry structure.