Skip to content

Materials MLIP candidate measurement v1

  • Stability: development
  • Revision: v1-development
  • Scope: equal-atom OMat24 candidate environments measured under the frozen Nequix-MP-1 starting potential before candidate post-training

This convention projects the released MLIP phonon candidate bank into Anarres environment profiles. The native candidate-bank artifact remains authoritative. Profiles contain measurements made before candidate training and exclude development or prospective phonon outcomes.

The measurement contract uses one deterministic starting-model evaluation per retained structure. Every candidate contains exactly 50,000 processed atoms and at most one configuration from each OMat24 parent. Candidate membership is identified by the SHA-256 digest of its ordered native structure identities.

signal_density_per_atom

Loss-weighted starting-model residual mass divided by processed atoms. For one structure with atom count $n$, mean force-vector residual $r_f$, and mean absolute stress residual $r_s$, the native mass is

$$ A = 20 n r_f + 5 r_s. $$

The candidate coordinate is $\sum A / \sum n$. Its unit is loss_weighted_residual_per_atom, its estimator is the atom-normalized sum over retained structures, and its direction is contextual. The coefficients are part of the frozen measurement contract and do not make force and stress residuals physically interchangeable.

effective_signal_coverage

Normalized effective sample size of the structure-level signal masses. For $m$ retained structures with masses $A_i$,

$$ H = \frac{(\sum_i A_i)^2}{m\sum_i A_i^2}. $$

The value lies between zero and one, with zero defined when total mass is zero. Its unit is fraction, its estimator is the normalized effective sample size, and its direction is contextual until downstream candidate outcomes validate a decision rule.

chemical_system_entropy

Normalized Shannon entropy over the retained structures' element-set labels. The unweighted label frequency is used, and the entropy is divided by the log of the number of observed labels. The value is zero for one observed label and lies between zero and one otherwise. Its unit is fraction, and its direction is contextual.

mean_target_affinity

Arithmetic mean of each retained structure's maximum element-set Jaccard similarity to the 16 checkpoint-qualification targets. Only element identities from the qualification split enter this measurement. Development and prospective phonon outcomes did not enter its construction. Its unit is fraction, and its direction is contextual.

neighbor_edges_per_atom

Total Nequix neighbor edges divided by processed atoms for the candidate. This is a model- and cutoff-specific compute proxy, not a universal hardware-cost estimate. Its unit is neighbor_edges_per_atom, and lower values indicate less graph-processing work under the frozen model-input contract.

processed_atom_budget

Total atom instances in the candidate environment. Its unit is processed_atoms, and all profiles in the released draft equal 50,000. This coordinate records the matched primary data budget. It does not imply that neighbor-edge cost is matched.

Qualification and edge cases

The adapter requires the pinned OMat24 source qualification, Nequix-MP-1 model hash, frozen Alexandria target roles, overlap audit, complete 80,000-record measurement validation, and the outcome-blind candidate-bank artifact. Missing or nonfinite measurements fail export. Zero measurements are preserved.

The published bank remains a measurement-only draft. Its candidates are valid for comparing pre-training coordinates, while the final training panel remains unfrozen because neighbor-edge cost varies across candidates. This condition is reported as a warning and never converted into a downstream outcome.