Search

Process Mining offers a powerful framework for uncovering, analyzing, and optimizing real-world business processes. Petri nets provide a versatile means of modeling process behavior. However, traditional methods often struggle to effectively compare complex Petri nets, hindering their potential for process enhancement. To address this challenge, we introduce PetriNet2Vec, an unsupervised methodology inspired by Doc2Vec. This approach converts Petri nets into embedding vectors, facilitating the comparison, clustering, and classification of process models. We validated our approach using the PDC Dataset, comprising 96 diverse Petri net models. The results demonstrate that PetriNet2Vec effectively captures the structural properties of process models, enabling accurate process classification and efficient process retrieval. Specifically, our findings highlight the utility of the learned embeddings in two key downstream tasks: process classification and process retrieval. In process classification, the embeddings allowed for accurate categorization of process models based on their structural properties. In process retrieval, the embeddings enabled efficient retrieval of similar process models using cosine distance. These results demonstrate the potential of PetriNet2Vec to significantly enhance process mining capabilities.

We demonstrate the miniaturization of perpendicularly magnetized superparamagnetic tunnel junctions (SMTJs) down to 50 nm in diameter. We experimentally show stochastic reversals in those junctions, with tunable mean dwell times down to a few nanoseconds through applied magnetic field and voltage. The mean dwell times measured at negligible bias voltage agree with our simulations based on Langer's theory. We shed light on an Arrhenius prefactor $\tau_0$ of a few femtoseconds, implying that the rates of thermally-activated magnetic transitions exceed the GHz-to-THz limitation of macrospin models, whereby $\tau_0\sim1$ ns. We explain the small prefactor values by a Meyer-Neldel compensation phenomenon, where the prefactor exhibits a large entropic contribution with an exponential dependence on the activation energy. These findings pave the way towards the development of ultrafast, low-power unconventional computing schemes operating by leveraging thermal noise in perpendicular SMTJs, which are scalable below 20 nm., Comment: 7 pages, 3 figures

Spintronic devices have recently attracted a lot of attention in the field of unconventional computing due to their non-volatility for short and long term memory, non-linear fast response and relatively small footprint. Here we report how voltage driven magnetization dynamics of dual free layer perpendicular magnetic tunnel junctions enable to emulate spiking neurons in hardware. The output spiking rate was controlled by varying the dc bias voltage across the device. The field-free operation of this two terminal device and its robustness against an externally applied magnetic field make it a suitable candidate to mimic neuron response in a dense Neural Network (NN). The small energy consumption of the device (4-16 pJ/spike) and its scalability are important benefits for embedded applications. This compact perpendicular magnetic tunnel junction structure could finally bring spiking neural networks (SNN) to sub-100nm size elements.

Dialogue systems need to deal with the unpredictability of user intents to track dialogue state and the heterogeneity of slots to understand user preferences. In this paper we investigate the hypothesis that solving these challenges as one unified model will allow the transfer of parameter support data across the different tasks. The proposed principled model is based on a Transformer encoder, trained on multiple tasks, and leveraged by a rich input that conditions the model on the target inferences. Conditioning the Transformer encoder on multiple target inferences over the same corpus, i.e., intent and multiple slot types, allows learning richer language interactions than a single-task model would be able to. In fact, experimental results demonstrate that conditioning the model on an increasing number of dialogue inference tasks leads to improved results: on the MultiWOZ dataset, the joint intent and slot detection can be improved by 3.2\% by conditioning on intent, 10.8\% by conditioning on slot and 14.4\% by conditioning on both intent and slots. Moreover, on real conversations with Farfetch costumers, the proposed conditioned BERT can achieve high joint-goal and intent detection performance throughout a dialogue.

Magnetic skyrmions are topological spin textures which are envisioned as nanometre scale information carriers in magnetic memory and logic devices. The recent demonstration of room temperature stabilization of skyrmions and their current induced manipulation in industry compatible ultrathin films were first steps towards the realisation of such devices. However, important challenges remain regarding the electrical detection and the low-power nucleation of skyrmions, which are required for the read and write operations. Here, we demonstrate, using operando magnetic microscopy experiments, the electrical detection of a single magnetic skyrmion in a magnetic tunnel junction (MTJ) and its nucleation and annihilation by gate voltage via voltage control of magnetic anisotropy. The nucleated skyrmion can be further manipulated by both gate voltage and external magnetic field, leading to tunable intermediate resistance states. Our results unambiguously demonstrate the readout and voltage controlled write operations in a single MTJ device, which is a major milestone for low power skyrmion based technologies.

Magnetic tunnel junctions (MTJs) are elementary units of magnetic memory devices. For high-speed and low-power data storage and processing applications, fast reversal by an ultrashort laser pulse is extremely important. We demonstrate optical switching of Tb/Comultilayer-based nanoscale MTJs by combining optical writing and electrical read-out methods. A 90 fs-long laser pulse switches the magnetization of the storage layer (SL). The change in magnetoresistance between the SL and a reference layer (RL) is probed electrically across the tunnel barrier. Single-shot switching is demonstrated by varying the cell diameter from 300 nm to 20 nm. The anisotropy, magnetostatic coupling, and switching probability exhibit cell-size dependence. By suitable association of laser fluence and magnetic field, successive commutation between high-resistance and low-resistance states is achieved. The switching dynamics in a continuous film is probed with the magneto-optical Kerr effect technique. Our experimental findings provide strong support for the growing interest in ultrafast spintronic devices., Comment: total pages 22, Total figure 8

Perpendicular shape anisotropy (PSA) offers a practical solution to downscale spin-transfer torque Magnetic Random-Access Memory (STT-MRAM) beyond the sub-20 nm technology node whilst retaining thermal stability of the storage layer magnetization. However, our understanding of the thermomagnetic behavior of PSA-STT-MRAM is often indirect, relying on magnetoresistance measurements and micromagnetic modelling. Here, the magnetism of a FeCoB / NiFe PSA-STT-MRAM nano-pillar is investigated using off-axis electron holography, providing spatially resolved magnetic information as a function of temperature, which has been previously inaccessible. Magnetic induction maps reveal the micromagnetic configuration of the NiFe storage layer (60 nm high, 20 nm diameter), confirming the PSA induced by its 3:1 aspect ratio. In-situ heating demonstrates that the PSA of the FeCoB / NiFe composite storage layer is maintained up to at least 250 degrees centigrade, and direct quantitative measurements reveal the very moderate decrease of magnetic induction with temperature. Hence, this study shows explicitly that PSA provides significant stability in STT-MRAM applications that require reliable performance over a range of operating temperatures.

We report on the experimental evidence of magnetic helicoidal dichroism, observed in the interaction of an extreme ultraviolet vortex beam carrying orbital angular momentum with a magnetic vortex. Numerical simulations based on classical electromagnetic theory show that this dichroism is based on the interference of light modes with different orbital angular momenta, which are populated after the interaction between the light phase chirality and the magnetic topology. This observation gives insight into the interplay between orbital angular momentum and magnetism, and sets the framework for the development of new analytical tools to investigate ultrafast magnetization dynamics., Comment: 7 pages, 4 figures

Since the first experimental observation of all-optical switching phenomena, intensive research has been focused on finding suitable magnetic systems that can be integrated as storage elements within spintronic devices and whose magnetization can be controlled through ultra-short single laser pulses. We report here atomistic spin simulations of all-optical switching in multilayered structures alternating n monolayers of Tb and m monolayers of Co. By using a two temperature model, we numerically calculate the thermal variation of the magnetization of each sublattice as well as the magnetization dynamics of [Tbn/Com] multilayers upon incidence of a single laser pulse. In particular, the condition to observe thermally-induced magnetization switching is investigated upon varying systematically both the composition of the sample (n,m) and the laser fluence. The samples with one monolayer of Tb as [Tb1/Co2] and [Tb1/Co3] are showing thermally induced magnetization switching above a fluence threshold. The reversal mechanism is mediated by the residual magnetization of the Tb lattice while the Co is fully demagnetized in agreement with the models developed for ferrimagnetic alloys. The switching is however not fully deterministic but the error rate can be tuned by the damping parameter. Increasing the number of monolayers the switching becomes completely stochastic. The intermixing at the Tb/Co interfaces appears to be a promising way to reduce the stochasticity. These results predict for the first time the possibility of TIMS in [Tb/Co] multilayers and suggest the occurrence of sub-picosecond magnetization reversal using single laser pulses., Comment: 9 pages, 4 figures