Altermagnets exhibit spin-split electronic structure without requiring spin-orbit coupling, but transport measurements generally mix intrinsic spin splitting with extrinsic scattering. We examine this identifiability problem for a two-dimensional $d$-wave altermagnet within a unified semiclassical framework spanning ballistic to diffusive transport. The spin-dependent Fermi-surface anisotropy produces a pronounced size effect, where vastly different longitudinal velocities cause the two spin channels to exhibit markedly different effective relaxation lengths within the same device geometry. However, the altermagnetic coupling $\alpha$ and the momentum relaxation time ${\tau }_0$ strongly compensate each other in longitudinal conductance, producing strong parameter correlation. To address this inverse problem, we formulate a physics-informed neural network (PINN) as a differentiable Boltzmann solver with contact-injection and current-continuity losses and a particle-conserving local distribution. Driven by sparse synthetic conductance spectra, this solver uses the Fermi-level dependence of transport to recover the coupled parameters under the tested conditions, including joint inversion at 1% Gaussian noise and single-parameter extraction at noise levels up to 10%. These results demonstrate parameter recovery by combining Fermi-level-dependent transport data with a physically constrained numerical solver within the model and parameter range studied.

Close

PrevNext

View figure in article

• Altermagnetism
• Magnetotransport
• Spintronics
• Magnetic systems
• Machine learning