LitXAlloy: A Multifaceted Alloy Framework

• LitXAlloy is a versatile alloy research framework with context-dependent meanings ranging from literature extraction benchmarks to experimental alloy characterization.
• It enables dense extraction of experimental processes through process-linked material objects and precise quantitative metrics like F1 scores.
• The framework supports advanced alloy design by integrating corrosion resistance analysis, beamline material evaluation, and computational screening of complex alloys.

LitXAlloy is a context-dependent label used across recent alloy and materials-informatics literature for several distinct but related objects. The published usage suggests that it is not a single standardized material designation, but rather a name applied to alloy-centered benchmarks, alloy compositions, and screening frameworks. In one major usage, LitXAlloy is the dense alloy subset of LitXBench for experiment extraction from scientific papers (Chong et al., 8 Apr 2026). In another, it denotes a corrosion-resistant, ultralight Mg–Li–(Al–Y–Zr) sheet alloy in the high-Li, single-phase bcc regime (Yan et al., 2019). The term is also used for Mg–Al–Li beampipe design studies, correlative alloy-anode characterization programs, and alloy-screening formulations that connect phase stability, precipitation, transport, and ductility to explicit quantitative metrics (Singh et al., 19 Jan 2026).

1. Terminological scope and usage

The literature supports three principal usages of LitXAlloy. First, it appears as a data-centric benchmark inside LitXBench, where the objective is extraction of complete experiments from alloy papers as structured objects rather than flat records (Chong et al., 8 Apr 2026). Second, it appears as a materials designation for a high-Li Mg-based alloy whose corrosion resistance is tied to a stratified surface film containing $\mathrm{Li_2CO_3}$ and $\mathrm{Mg(OH)_2}$ (Yan et al., 2019). Third, it functions as a broader alloy-design label in studies that optimize radiation length and stiffness, track alloying-induced phase transformations operando, or screen chemically complex systems by first-principles or machine-learning-based criteria (Singh et al., 19 Jan 2026).

This multiplicity of usage is significant because the common thread is methodological rather than purely compositional. Across the cited works, LitXAlloy is associated with dense experiment representation, cross-scale alloy characterization, and composition–process–structure–property coupling. A plausible implication is that the term is best understood as an alloy-centered research framework whose meaning depends on the surrounding domain: literature extraction, corrosion science, structural alloy design, electrochemical alloying, or computational screening.

2. LitXAlloy as a benchmark for extracting alloy experiments

Within LitXBench, LitXAlloy is the alloy-focused evaluation set designed to benchmark extraction of complete experiments—materials, synthesis processes, and all reported measurements—from full scientific papers (Chong et al., 8 Apr 2026). The task is formalized as returning tuples $(m_i, p_i, x_i)$ for every material synthesized in a paper. The dataset contains 19 alloy papers and 1426 total experimental and experimentally-derived measurements, with 101 target materials and 68 unique compositions. The benchmark explicitly emphasizes that “materials are not compositions”: 8 papers contain duplicate compositions, with 12 duplicates overall, and 26 materials across 6 papers are derived from other materials.

Its data model is code-first. Entries are stored as Python objects rather than CSV or JSON, with core classes including Experiment, RawMaterial, Material, ProcessEvent, Measurement, CompositionMeasurement, GlobalLatticeParam, and Configuration. Canonicalization is performed with enums such as AlloyMeasurementKind, PhaseMeasurementKind, ProcessKind, and MeasurementMethod; units are standardized with Pint; compositions are normalized with Pymatgen Composition; and helper functions are provided for balance notation and weight-fraction additions. This representation is intended to improve auditability, preserve provenance through normalize(val, val_in_paper) mappings, and enable compile-time and run-time validation.

The benchmark also encodes process lineage explicitly. Materials declare their synthesis history using compact arrow notation, forming a directed acyclic graph. Programmatic invariants enforce structural consistency, including the absence of graph cycles, mandatory casting steps after melting events, valid compositions, unit-complete quantities, and coverage of all raw materials and synthesis groups. This is a substantive departure from composition-centric extraction pipelines, because measurements are attached to process-defined materials rather than nominal formulas alone.

Evaluation uses category-level F1 scores for measurements, process, material count, and configuration, combined into the weighted score

$$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$

Extracted and target materials are paired with the Hungarian algorithm. Measurement matching accounts for kind, value and unit equivalence, qualifiers, temperature, and pressure; process order is scored by Levenshtein distance; and configuration scoring includes hierarchical parent–child consistency. On this benchmark, Gemini CLI (3.1 Pro Preview) reached $F1 = 0.80 \pm 0.04$, with Measurement 0.74, Process 0.84, Material 0.98, and Configuration 0.68, while frontier models outperformed composition-centric multi-turn pipelines by up to 0.37 F1 (Chong et al., 8 Apr 2026).

In this usage, LitXAlloy is not a physical alloy but a dense, auditable benchmark for alloy knowledge extraction. Its main conceptual contribution is the shift from composition-linked records to process-linked material objects, which aligns the representation of literature data with how alloy properties are actually generated and reported.

3. LitXAlloy as a corrosion-resistant Mg–Li–(Al–Y–Zr) sheet alloy

In corrosion science, LitXAlloy denotes a high-Li Mg–Li–(Al–Y–Zr) sheet alloy deliberately positioned in the single-phase $\beta$ bcc regime (Yan et al., 2019). Its nominal bulk composition is Mg (balance) – 30.3 at.% Li – 2.34 at.% Al – 0.128 at.% Y – 0.039 at.% Zr. The alloy was prepared by gravity casting, followed by homogenisation, extrusion, heat treatment and water quenching, artificial ageing, and cold rolling. Surface films were formed by immersion in quiescent 0.1 M NaCl for 24 h, followed by ethanol rinsing and 7 days of air exposure.

The surface film was characterized by grazing incidence X-ray diffraction, SAED, STEM–EDXS, and EELS. GIXRD identified $\mathrm{Mg(OH)_2}$ and a weaker signal from $\mathrm{Li_2CO_3}$ at the most surface-sensitive angle. At larger incidence angles, intense signals from hcp $\alpha$-Mg appeared in addition to bcc $\beta$-Li, implying enrichment of $\mathrm{Mg(OH)_2}$0-Mg near the surface despite the bcc bulk. Cross-sectional TEM resolved four sections: the protective Pt deposition layer, a porous outer layer, a denser inner layer, and the matrix. Bright-field TEM showed needle- or platelet-shaped corrosion products in the outer layer, and SAED confirmed hcp reflections beneath the film.

EELS demonstrated that Li is distributed throughout the film, with maximum Li intensity in the outer layer. The layered film is therefore interpreted as a porous, Li-rich outer layer dominated by $\mathrm{Mg(OH)_2}$1, sitting on a denser, Mg-enriched inner layer. The outer layer extends through the first $\mathrm{Mg(OH)_2}$2 nm from the surface, while the inner layer reaches to $\mathrm{Mg(OH)_2}$3 nm. Total film thickness is thus on the order of $\mathrm{Mg(OH)_2}$4 nm. Mg is more intense in the inner layer, while Li remains above the matrix baseline throughout the film.

A central mechanistic result is the detection of hcp $\mathrm{Mg(OH)_2}$5-Mg immediately beneath the film. The proposed explanation is selective dissolution of Li during immersion, reducing the local Li wt.% from $\mathrm{Mg(OH)_2}$6 wt.% below the $\mathrm{Mg(OH)_2}$7 wt.% threshold required to stabilize the single-phase bcc matrix. This enables local nucleation of $\mathrm{Mg(OH)_2}$8-Mg near the surface. Such $\mathrm{Mg(OH)_2}$9–$(m_i, p_i, x_i)$0 heterogeneity can promote galvanic coupling, but the stratified film mitigates further dissolution.

The corrosion resistance of the alloy is therefore attributed to synergy between the high-Li bcc bulk and the surface film. The outer $(m_i, p_i, x_i)$1-rich layer acts as a passivating barrier despite its porosity, and the denser inner Mg-enriched layer contributes additional barrier function and mechanical continuity. Al is detected across the inner layer and is consistent with previously reported Al-containing Li-LDH in the same alloy system, while weak $(m_i, p_i, x_i)$2 dispersoid reflections are observed in GIXRD. Y- and Zr-containing products, if present, remain below the detection threshold (Yan et al., 2019).

4. LitXAlloy as a Mg–Al–Li beampipe materials program

In accelerator materials design, LitXAlloy is conceived as a next-generation beampipe material that simultaneously maximizes radiation length while retaining sufficient stiffness and robust vacuum compatibility (Singh et al., 19 Jan 2026). The governing figure of merit is

$(m_i, p_i, x_i)$3

where $(m_i, p_i, x_i)$4 is radiation length and $(m_i, p_i, x_i)$5 is Young’s modulus. Conventional benchmarks in the study are Stainless steel 304 with $(m_i, p_i, x_i)$6 m, $(m_i, p_i, x_i)$7 GPa, $(m_i, p_i, x_i)$8; Aluminum with $(m_i, p_i, x_i)$9 m, $$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$0 GPa, $$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$1; and an Al–Ti–V reference alloy with $$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$2 m, $$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$3 GPa, $$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$4.

Five ternary Mg–Al–Li alloys were screened by Thermo-Calc TC-Python and DFT: two Al-rich alloys, A1 and A2, and three Mg-rich alloys, M1, M2, and M3. Thermodynamic stability, density, liquidus temperature, and phases were evaluated using Latin hypercube sampling within TC-Python, using TCAL8 for A1 and A2 and TCMG6 for M1, M2, and M3. Elastic properties were obtained from VASP with PBE-GGA, PAW pseudopotentials, a 520 eV cutoff, and Voigt–Reuss–Hill reduction of the stiffness tensor.

Alloy	Composition (at%)	$$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$5, $$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$6, $$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$7, $$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$8
A1	Al61.5Li10.8Mg27.7	2108 kg/m$$F1_{\mathrm{overall}} = 0.5\,F1_{\mathrm{Meas}} + 0.2\,F1_{\mathrm{Proc}} + 0.15\,F1_{\mathrm{Mat}} + 0.15\,F1_{\mathrm{Config}}.$$9; 0.1178 m; 99.66 GPa; 0.5460
A2	Al66Li19.4Mg14.6	2035 kg/m$F1 = 0.80 \pm 0.04$0; 0.1240 m; 117.43 GPa; 0.6072
M1	Al23.9Li29.3Mg46.8	1559 kg/m$F1 = 0.80 \pm 0.04$1; 0.1703 m; 27.538 GPa; 0.5140
M2	Al19Li20.6Mg60.4	1619 kg/m$F1 = 0.80 \pm 0.04$2; 0.1592 m; 66.07 GPa; 0.6436
M3	Al39.8Li20.1Mg40.1	1795 kg/m$F1 = 0.80 \pm 0.04$3; 0.1428 m; 104.64 GPa; 0.6729

These data show that all five candidates improve on aluminum in $F1 = 0.80 \pm 0.04$4, with gains of +48% for A1, +64% for A2, +39% for M1, +74% for M2, and +82% for M3. Relative to stainless steel, the candidates deliver approximately 5.0–6.6× higher $F1 = 0.80 \pm 0.04$5. M3 is identified as the best near-term choice, with the highest $F1 = 0.80 \pm 0.04$6, while A2 is presented as an Al-rich alternative prioritizing manufacturability and joining. M2 is attractive when very high radiation length is prioritized but stiffness demands remain moderate.

The underlying phase equilibria are multiphase. At 673 K, the ternary system spans HCP_A3, Al1Li1, Al12Mg17_A12, and Li-rich BCC_B2. M1 is dominated by HCP_A3 near the liquidus with increasing AlLi on cooling; M2 forms primary HCP_A3 followed by AlLi and Al12Mg17_A12; and M3 follows a more complex path involving AlLiMg_T, Al12Mg17_A12, AlLi, and a terminal eutectic. The study interprets these stable multiphase microstructures as central to the LitXAlloy performance envelope.

Practical constraints remain explicit. The work notes the need for validation of corrosion behavior, outgassing, bake-out stability, irradiation response, fracture toughness, fatigue, creep, and weld metallurgy. Reported elastic data are 0 K DFT VRH averages, and detailed uncertainty estimates are not provided. Even so, the program establishes LitXAlloy in this context as a quantitatively specified family of low-$F1 = 0.80 \pm 0.04$7, high-$F1 = 0.80 \pm 0.04$8, comparatively stiff ternaries for beamline applications (Singh et al., 19 Jan 2026).

5. Cross-scale correlative and screening frameworks

A further usage of LitXAlloy appears in cross-scale alloying studies that combine operando probes, cryogenic microscopy, and mechanistic transport or phase models. In a Pt-based alloying anode study, operando hard X-ray nanoprobe XRD and XRF in a liquid electrochemical nanocell were combined with cryo-STEM/EELS and cryo-APT to resolve lithiation and delithiation across length scales (Mulcahy et al., 20 Nov 2025). Lithiation of polycrystalline fcc Pt proceeds by initial formation of $F1 = 0.80 \pm 0.04$9, followed by evolution to a stable LiPt intermetallic with extended cycling. Operando XRD shows discrete $\beta$0 reflections on lithiation; these disappear on delithiation, while a broad LiPt (0001) peak emerges and shifts from $\beta$1 $\beta$2 to $\beta$3 $\beta$4, corresponding to $\beta$5 $\beta$6 to $\beta$7 $\beta$8. Cryo-STEM/EELS shows transition of the SEI from an unstable carbonate-rich composition to a LiF-dominated interphase, and cryo-APT identifies an interfacial Li-flux-limited heterogeneous zone above a diffusion-controlled homogeneous LiPt bulk. The study explicitly frames these results as a broadly applicable correlative framework for LitXAlloy-style alloy-electrode design.

A related screening logic appears in mixed-halide solid-state electrolytes, where the system $\beta$9 is treated as an alloyed ionic conductor (Tisi et al., 19 Nov 2025). Both monoclinic $\mathrm{Mg(OH)_2}$0 and trigonal $\mathrm{Mg(OH)_2}$1 polymorphs were simulated from $\mathrm{Mg(OH)_2}$2 to $\mathrm{Mg(OH)_2}$3. The relative stability crossover occurs near $\mathrm{Mg(OH)_2}$4, mixing enthalpies are small, and ideal-solution configurational entropy at $\mathrm{Mg(OH)_2}$5 is 10.8 meV/atom at 300 K. Hybrid MC/MD showed that the local composition of each $\mathrm{Mg(OH)_2}$6 octahedron follows the binomial distribution expected for random occupation, while conductivity calculations demonstrated a compensation between chemistry and volume: Cl substitution contracts the lattice and tightens migration bottlenecks, but chemically lowers Li migration barriers. At fixed volume, a +5% expansion roughly doubles conductivity; under NPT conditions, these effects partially cancel.

The same general principle—correlative characterization linked to explicit design rules—also appears in first-principles work on Mg–Li–Al light alloys (Orhan et al., 2022). There, random solid solutions are ruled out because VCA formation enthalpies are positive, in the range 10–100 meV atom$\mathrm{Mg(OH)_2}$7, and ordered BCC-derived intermetallics become the design focus. The study identifies MgLi$\mathrm{Mg(OH)_2}$8Al in $\mathrm{Mg(OH)_2}$9 and $\mathrm{Li_2CO_3}$0-43m variants as promising, with predicted $\mathrm{Li_2CO_3}$1–1.3 GPa at $\mathrm{Li_2CO_3}$2 g cm$\mathrm{Li_2CO_3}$3, while D0$\mathrm{Li_2CO_3}$4–Mg$\mathrm{Li_2CO_3}$5Al is notable as a precipitation-strengthening phase rather than a maximally hard intrinsic phase. The rule of mixtures is used to upscale phase properties to heterogeneous architectures. In this usage, LitXAlloy denotes a design space governed by stability, density, and architecture-level strength rather than a single nominal composition.

6. Precipitation, site occupancy, and phase equilibria in Al–Li-based systems

Several alloy-metallurgy studies linked to LitXAlloy focus on Al–Li-based precipitation systems. In an Al–Cu–Li diffusion-couple study, Mg concentration was varied continuously from $\mathrm{Li_2CO_3}$6 to 0.35 wt% across $\mathrm{Li_2CO_3}$7 mm using a linear-friction-welded couple, and precipitation kinetics were monitored by in-situ SAXS and microhardness during ageing at 155$\mathrm{Li_2CO_3}$8C (Gumbmann et al., 2015). A clear transition in kinetics occurs around 0.1–0.2 wt% Mg. The Mg-free alloy reaches maximum hardness after $\mathrm{Li_2CO_3}$9 h, whereas the 0.35 wt% Mg alloy peaks after $\alpha$0 h, approximately three times faster. Mg-bearing regions exceed Mg-free regions by $\alpha$1 HV at the plateau and by up to $\alpha$2 HV at intermediate times around 12 h. SAXS-derived relative volume fraction, estimated from the integrated intensity $\alpha$3, correlates strongly with hardness during the first $\alpha$4 h. The interpretation is that Mg promotes precipitation of the T1-$\alpha$5 phase, with saturation behavior above $\alpha$6 wt% Mg.

A complementary atomic-scale result was obtained for $\alpha$7 precipitates in Al–Mg–Li (Gault et al., 2015). Atom probe microscopy and DFT showed that $\alpha$8 is an ordered, coherent $\alpha$9 Al$\beta$0(Li,Mg) phase, and that Mg occupies the Li-occupied $\beta$1 sublattice rather than the Al-rich $\beta$2 sublattices. Experimentally, Mg concentration in $\beta$3 is approximately the same as in the matrix, but species-specific spatial distribution maps show Mg peaks coincident with Li peaks inside $\beta$4. DFT supports this site preference quantitatively: $\beta$5 eV, $\beta$6 eV, and $\beta$7 eV. This reconciles weak concentration partitioning with strong site-specific partitioning.

At the phase-diagram level, the Al-rich corner of Al–Li–Cu has been computed from first principles and statistical mechanics by Liu, Wróbel, and Llorca (Liu et al., 2022). Ground-state phases in the Al-rich region are $\beta$8-Al, $\beta$9 ($\mathrm{Mg(OH)_2}$00), $\mathrm{Mg(OH)_2}$01 ($\mathrm{Mg(OH)_2}$02), $\mathrm{Mg(OH)_2}$03 (AlLi), and T1, with T1 predicted as hexagonal $\mathrm{Mg(OH)_2}$04. Metastable phases include $\mathrm{Mg(OH)_2}$05 ($\mathrm{Mg(OH)_2}$06), T1′ ($\mathrm{Mg(OH)_2}$07), and $\mathrm{Mg(OH)_2}$08. A key thermodynamic nuance is that $\mathrm{Mg(OH)_2}$09, while on the 0 K hull, becomes metastable already at very low temperature when configurational entropy is included: its distance from the $\mathrm{Mg(OH)_2}$10–$\mathrm{Mg(OH)_2}$11 tie-line changes from $\mathrm{Mg(OH)_2}$12 meV/atom at 0 K to $\mathrm{Mg(OH)_2}$13 meV/atom at 100 K. The work also emphasizes that $\mathrm{Mg(OH)_2}$14 replaces $\mathrm{Mg(OH)_2}$15 as the stable phase at approximately 550 K owing to vibrational entropy. Together, these studies define a LitXAlloy-related picture in which precipitation kinetics, sublattice occupancy, and metastability are all composition- and temperature-sensitive.

7. Electronic-structure metrics for superalloys and refractory alloys

In nickel superalloys, LitXAlloy is associated with composition–energy design of $\mathrm{Mg(OH)_2}$16-like precipitates. DFT combined with multiple scattering theory has been used to compute $\mathrm{Mg(OH)_2}$17 and $\mathrm{Mg(OH)_2}$18 for $\mathrm{Mg(OH)_2}$19 with $\mathrm{Mg(OH)_2}$20 Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, or W in $\mathrm{Mg(OH)_2}$21, $\mathrm{Mg(OH)_2}$22, and $\mathrm{Mg(OH)_2}$23 phases (Zarkevich et al., 2023). The cubic $\mathrm{Mg(OH)_2}$24 phase corresponds to $\mathrm{Mg(OH)_2}$25, while $\mathrm{Mg(OH)_2}$26 and $\mathrm{Mg(OH)_2}$27 represent hexagonal stackings associated with extrinsic and intrinsic stacking faults. Group IV solutes stabilize $\mathrm{Mg(OH)_2}$28 at sufficiently high $\mathrm{Mg(OH)_2}$29, while group V solutes stabilize $\mathrm{Mg(OH)_2}$30. For example, $\mathrm{Mg(OH)_2}$31 is predicted to undergo an $\mathrm{Mg(OH)_2}$32 change at $\mathrm{Mg(OH)_2}$33. Terminal compounds such as $\mathrm{Mg(OH)_2}$34 and $\mathrm{Mg(OH)_2}$35 have high bulk moduli, with $\mathrm{Mg(OH)_2}$36 GPa for $\mathrm{Mg(OH)_2}$37 $\mathrm{Mg(OH)_2}$38, but positive formation energies in some stackings. The design implication is that local composition can tune stacking-fault stabilization, lattice misfit, and stiffness in $\mathrm{Mg(OH)_2}$39.

A distinct but related design metric appears in bcc refractory multi-principal-element alloys, where the local–lattice distortion (LLD) has been proposed as a ductility predictor (Singh et al., 2022). The metric combines DFT-relaxed atomic displacements with a weighted valence-electron-count term,

$\mathrm{Mg(OH)_2}$40

with $\mathrm{Mg(OH)_2}$41 for the bcc window $\mathrm{Mg(OH)_2}$42. A single threshold separates ductile from brittle behavior: $\mathrm{Mg(OH)_2}$43 is classified as ductile, and $\mathrm{Mg(OH)_2}$44 as brittle. The metric was validated against tensile-elongation data and indentation crack formation. Representative values include NbTaTi with $\mathrm{Mg(OH)_2}$45, NbTaTiV with $\mathrm{Mg(OH)_2}$46, NbTaV with $\mathrm{Mg(OH)_2}$47, NbTaMoW with $\mathrm{Mg(OH)_2}$48, and $\mathrm{Mg(OH)_2}$49 with $\mathrm{Mg(OH)_2}$50. The electronic interpretation is that electronegativity-driven charge redistribution governs the relaxed displacement field and therefore the intrinsic ductile-to-brittle tendency.

Taken together, these studies place LitXAlloy within a broader alloy-design vocabulary grounded in explicit energetic, structural, and electronically informed metrics. Whether the target is $\mathrm{Mg(OH)_2}$51 fault stabilization, bcc refractory ductility, T1 precipitation kinetics, or beampipe radiation transparency, the recurring pattern is the use of quantitatively defined descriptors to connect local chemistry to macroscopic alloy performance.