arXiv subjects

Edward Schwartz

Publications and source records attributed to Edward Schwartz.

11 recordsLinked to original sources

arXiv2026

Decaf: Improving Neural Decompilation with Automatic Feedback and Search

Decompilers are useful tools used in reverse engineering to understand compiled source code. Reconstructing source code from compiled binaries is a challenging task, because high-level syntax, identifiers, and custom data types are generally lost as the compiler translates human-readable code to low-level machine code. Deterministic decompilers are useful tools for binary analysis, but can struggle to infer idiomatic syntax and identifier names. Generative AI models are a natural fit for reconstructing high-level syntax, identifiers, and types, but they can still suffer by hallucinating improper programming constructs and semantics. Instead of attempting to improve neural decompilers with more data and more training, we argue that compiler feedback can be used to dramatically improve the semantic correctness of neural decompiler outputs via search. Our system, Decaf (DECompilation with Automated Feedback), raises the neural decompilation rate from 26.0% on ExeBench to 83.9% on the Real -O2 split without sacrificing similarity to the original source code. We also find our automatic feedback methodology is highly effective for improving weaker neural decompilation models.

cs.SE↗

arXiv2025

Thermomagnonic Torques in Insulating Altermagnets

We develop a symmetry-controlled theory of anisotropic thermomagnonic torques in insulating altermagnets. We identify a spin-splitter magnonic torque linked to thermally generated, sublattice-odd spin currents and an anisotropic entropic torque dictated by crystal symmetry. These torques produce anisotropic magnetic-texture responses to temperature gradients. In particular, thermally generated spin currents induce domain-wall precession, which reduces domain-wall velocities for selected gradient directions. We also predict an anisotropic skyrmion Hall response, with symmetry-selected directions enabling fast skyrmion motion with strongly suppressed transverse deflection. Our results reveal experimentally testable symmetry fingerprints of insulating altermagnets and extend more broadly to anisotropic magnets with exchange-driven magnon spin splitting.

cond-mat.mes-hall↗

arXiv2025

Effect of Magnetic Anisotropy and Gradient-Induced Dzyaloshinskii-Moriya Interaction on the Formation of Magnetic Skyrmions

Topological spin textures (e.g. skyrmions) can be stabilized by interfacial Dzyaloshinskii-Moriya interaction (DMI) in the magnetic multilayer, which has been intensively studied. Recently, Bloch-type magnetic skyrmions stabilized by composition gradient-induced DMI (g-DMI) have been observed in 10-nm thick CoPt single layer. However, magnetic anisotropy in gradient-composition engineered CoPt (g-CoPt) films is highly sensitive to both the relative Co/Pt composition and the film thickness, leading to a complex interplay with g-DMI. The stability of skyrmions under the combined influence of magnetic anisotropy and g-DMI is crucial yet remains poorly understood. Here, we conduct a systematic study on the characteristics of magnetic skyrmions as a function of gradient polarity and effective gradient (defined as gradient/thickness) in g-CoPt single layers (thickness of 10-30 nm) using magnetic force microscopy (MFM), bulk magnetometry, and topological Hall effect measurements. Brillouin light scattering spectroscopy confirms that both the sign and magnitude of g-DMI depend on the polarity and amplitude of the composition gradient in g-CoPt films. MFM reveals that skyrmion size and density vary with g-CoPt film thickness, gradient polarity, and applied magnetic field. An increased skyrmion density is observed in samples exhibiting higher magnetic anisotropy, in agreement with micromagnetic simulations and energy barrier calculations.

cond-mat.mtrl-sci↗

arXiv2025

Spin-Transfer Torque in Altermagnets with Magnetic Textures

We predict the existence of anisotropic spin-transfer torque effect in textured altermagnets. To this end, we generalize the Zhang-Li torque to incorporate the symmetry associated with prototypical $d$-wave altermagnets and identify the spin-splitter adiabatic and nonadiabatic torques. Applying our results to domain wall dynamics induced by spin-transfer torque, we find that, in certain regimes, the spin-splitter adiabatic torque can induce domain wall precession, significantly slowing down domain wall motion. The response of the domain wall also becomes anisotropic, reflecting the $d$-wave symmetry of the altermagnet. Furthermore, we observe that the spin-splitter adiabatic torque modifies skyrmion dynamics, inducing anisotropic skyrmion Hall effect. The above phenomena can serve as a hallmark of altermagnetism in textured magnets, distinguishing it from the behavior of ordinary antiferromagnets.

cond-mat.mes-hall↗

arXiv2024

Spin-orbit torque in a three-fold-symmetric bilayer and its effect on magnetization dynamics

Field-free switching of perpendicular magnetization has been observed in an epitaxial L1$_1$-ordered CoPt/CuPt bilayer and attributed to spin-orbit torque (SOT) arising from the crystallographic $3m$ point group of the interface. Using a first-principles nonequilibrium Green's function formalism combined with the Anderson disorder model, we calculate the angular dependence of the SOT in a CoPt/CuPt bilayer and find that the magnitude of the $3m$ SOT is about 20% of the conventional dampinglike SOT. We further study the magnetization dynamics in perpendicularly magnetized films in the presence of $3m$ SOT and Dzyaloshinskii-Moriya interaction, using the equations of motion for domain wall dynamics and micromagnetic simulations. We find that for systems with strong interfacial DMI characterized by the Néel character of domain walls, a very large current density is required to achieve deterministic switching because reorientation of the magnetization inside the domain wall is necessary to induce the switching asymmetry. For thicker films with relatively weak interfacial DMI and the Bloch character of domain walls the deterministic switching with much smaller currents is possible, which agrees with recent experimental findings.

cond-mat.mtrl-sci↗

arXiv2024

Room Temperature Magnetic Skyrmions in Gradient-Composition Engineered CoPt Single Layers

Topologically protected magnetic skyrmions in magnetic materials are stabilized by interfacial or bulk Dzyaloshinskii-Moriya interaction (DMI). Interfacial DMI decays with increase of the magnetic layer thickness in just a few nanometers and bulk DMI typically stabilizes magnetic skyrmions at low temperatures. Consequently, more flexibility in manipulation of DMI is required for utilizing nanoscale skyrmions in energy efficient memory and logic devices at room temperature (RT). Here, we demonstrate the observation of RT skyrmions stabilized by gradient DMI (g-DMI) in composition gradient engineered CoPt single layer films by employing topological Hall effect, magnetic force microscopy, and nitrogen vacancy scanning magnetometry. Skyrmions remain stable at a wide range of applied magnetic fields and are confirmed to be nearly Bloch-type from micromagnetic simulation and analytical magnetization reconstruction. Furthermore, we observe skyrmion pairs which may be explained by skyrmion-antiskyrmion interactions. Our findings expand the family of magnetic materials hosting RT magnetic skyrmions by tuning g-DMI via gradient polarity and choice of magnetic elements.

cond-mat.mes-hall↗

arXiv2024

Nonreciprocal superfluidlike topological spin transport

We study superfluidlike spin transport facilitated by thermal diffusion of magnetic domain walls, where the positive and negative chiralities of domain walls act as opposite topological charges. The topological charge conservation leads to algebraic decay of spin current carried by domain walls, allowing for the transport of spin over extended distances. We demonstrate that the presence of the Dzyaloshinskii-Moriya interaction can lead to nonreciprocity in spin flow, thus effectively realizing a spin ratchet. In one scenario, the nonreciprocity arises due to diode-like behavior where the nucleation of domain walls is governed by thermal activation for one direction of spin current and by viscous injection for the other direction of spin current. We confirm our predictions by micromagnetic simulations of domain walls in TmIG nanowire.

cond-mat.mes-hall↗

arXiv2023

Magnetic Antiskyrmions in Two-Dimensional van der Waals Magnets Engineered by Layer Stacking

Magnetic skyrmions and antiskyrmions are topologically protected quasiparticles exhibiting a whirling spin texture in real space. Antiskyrmions offer some advantages over skyrmions as they are expected to have higher stability and can be electrically driven with no transverse motion. However, unlike the widely investigated skyrmions, antiskyrmions are rarely observed due to the required anisotropic Dzyaloshinskii-Moriya interaction (DMI). Here we propose to exploit the recently demonstrated van der Waals (vdW) assembly of two-dimensional (2D) materials that breaks inversion symmetry and creates conditions for anisotropic DMI. Using a 2D vdW magnet CrI${}_3$ as an example, we demonstrate, based on density functional theory (DFT) calculations, that this strategy is a promising platform to realize antiskyrmions. Polar layer stacking of two centrosymmetric magnetic monolayers of CrI${}_3$ efficiently lowers the symmetry, resulting in anisotropic DMI that supports antiskyrmions. The DMI is reversible by switching the ferroelectric polarization inherited from the polar layer stacking, offering the control of antiskyrmions by an electric field. Furthermore, we find that the magnetocrystalline anisotropy and DMI of CrI${}_3$ can be efficiently modulated by Mn doping, creating a possibility to control the size of antiskyrmions. Using atomistic spin dynamics simulations with the parameters obtained from our DFT calculations, we predict the formation of antiskyrmions in a Cr${}_{0.88}$Mn${}_{0.12}$I${}_3$ bilayer and switching their spin texture with polarization reversal. Our results open a new direction to generate and control magnetic antiskyrmions in 2D vdW magnetic systems.

cond-mat.mtrl-sci↗

arXiv2022

Spin Hall effect of vorticity

Using mapping between topological defects in an easy-plane magnet and electrical charges, we study interplay between vorticity and spin currents. We demonstrate that the flow of vorticity is accompanied by the transverse spin current generation -- an effect which can be termed as the spin Hall effect of vorticity. We study this effect across the BKT transition and establish the role of dissipation and spin non-conservation in the crossover from spin superfluidity to diffusive spin transport. Our results pave the way for low power computing devices relying on vorticity and spin flows.

cond-mat.mes-hall↗

arXiv2022

Superfluid Spin Transistor

We propose to use the Hall response of topological defects, such as merons and antimerons, to spin currents in 2D magnetic insulator with in-plane anisotropy for identification of the Berezinskii-Kosterlitz-Thouless (BKT) transition in a transistor-like geometry. Our numerical results relying on a combination of Monte Carlo and spin dynamics simulations show transition from spin superfluidity to conventional spin transport, accompanied by the universal jump of the spin stiffness and exponential growth of the transverse vorticity current. We propose a superfluid spin transistor in which the spin and vorticity currents are modulated by tuning the in-plane magnet across BKT transition, e.g., by changing the exchange interaction, magnetic anisotropy, or temperature.

cond-mat.mes-hall↗

arXiv2022

Learning to Superoptimize Real-world Programs

Program optimization is the process of modifying software to execute more efficiently. Superoptimizers attempt to find the optimal program by employing significantly more expensive search and constraint solving techniques. Generally, these methods do not scale well to programs in real development scenarios, and as a result, superoptimization has largely been confined to small-scale, domain-specific, and/or synthetic program benchmarks. In this paper, we propose a framework to learn to superoptimize real-world programs by using neural sequence-to-sequence models. We created a dataset consisting of over 25K real-world x86-64 assembly functions mined from open-source projects and propose an approach, Self Imitation Learning for Optimization (SILO) that is easy to implement and outperforms a standard policy gradient learning approach on our dataset. Our method, SILO, superoptimizes 5.9% of our test set when compared with the gcc version 10.3 compiler's aggressive optimization level -O3. We also report that SILO's rate of superoptimization on our test set is over five times that of a standard policy gradient approach and a model pre-trained on compiler optimization demonstration.

cs.LG↗