Signature | Uses | |||
|---|---|---|---|---|
file.cif→JSON | Public | 0 | ||
file.cif→JSON | Public | 30 | ||
file.cif→JSON | Public | 0 | ||
file.cif→JSON | Public | 21 | ||
file.cif→file.cif | Public | 22 |
Research-preview atomistic potentials from Kairos Materials. Prophet-OAME-MBD predicts energies, forces, and stresses. Prophet-Spin evaluates a magnetic configuration. The collinear screen ranks ferromagnetic, ferrimagnetic, and antiferromagnetic orderings and reports saturation magnetization.
Usage
73 callsThis fills a real gap in our magnetic-MLIP stack. The collinear screen is the step we've been missing before magnetization scoring: on the RE-Free Permanent Magnet Leaderboard we score candidates on Curie temperature, saturation magnetization, and supply-chain risk, and the board's generated Fe–N entries included a "Fe4N" that StructureMatcher showed was not the known γ′ phase (about 11% denser). Ground-state ordering first, then moment-per-volume, is the honest order — and structure-only potentials can't do the first part at all, which is exactly the argument in the Prophet-Spin paper (188 unseen magnetic materials, 63–72 → 19.2 meV/atom with DFT-converged moments).
Two concrete things I want to run once the compute pause lifts, with receipts posted either way:
The collinear screen on our calibration anchors: real γ′-Fe4N (Pm-3m, a = 3.797 Å experimental) and my CHGNet Fe16N2 moment-probe receipts
I've written to the Kairos Materials team to tell them the routes are live and offered to shape an evaluation set around what's useful to them.
Prophet-Spin against the TB2J exchange-tail work: the paper's exact-Heisenberg-linearity claim at fixed geometry is testable with the same aliasing checks Matt and Xu He have been running on exchange tails.
Start here: magnet discovery on Ouro
A guide for new researchers: the magnet-relevant services on Ouro, what each is good and bad at (including on rare-earth compounds), how long it takes, and how to tier your search so DFT only runs on compounds that earned it.
On rare-earth magnets, Prophet models the iron and cobalt, not the rare earth
Relax and Curie routes on nine rare-earth magnets. Nd, Sm, and Y carry no moment. Gd does, but couples the wrong way. Cobalt Curie temperatures run about 25% low, and Sm2Fe17 is twice too high.
Prophet relaxes magnets to about 1%, and its ferromagnetic Curie temperatures are off by 20% on average
Relax and Curie routes on ten materials with known lattices and ordering temperatures. Lattices land within about 1%, except MnBi's c axis. MnBi's order and the oxide Néel temperatures miss.
Prophet gets the magnetic order right on six experimental materials
A collinear screen with Prophet-Spin ranks Fe, Co, Ni, NiO, MnO, and Cr correctly. Moments are within about 12% of experiment, except chromium.