Search

While deep-learning models have demonstrated skillful El Ni\~no Southern Oscillation (ENSO) forecasts up to one year in advance, they are predominantly trained on climate model simulations that provide thousands of years of training data at the expense of introducing climate model biases. Simpler Linear Inverse Models (LIMs) trained on the much shorter observational record also make skillful ENSO predictions but do not capture predictable nonlinear processes. This motivates a hybrid approach, combining the LIMs modest data needs with a deep-learning non-Markovian correction of the LIM. For O(100 yr) datasets, our resulting Hybrid model is more skillful than the LIM while also exceeding the skill of a full deep-learning model. Additionally, while the most predictable ENSO events are still identified in advance by the LIM, they are better predicted by the Hybrid model, especially in the western tropical Pacific for leads beyond about 9 months, by capturing the subsequent asymmetric (warm versus cold phases) evolution of ENSO.

Image Captioning is an important Language and Vision task that finds application in a variety of contexts, ranging from healthcare to autonomous vehicles. As many real-world applications rely on devices with limited resources, much effort in the field was put into the development of lighter and faster models. However, much of the current optimizations focus on the Transformer architecture in contrast to the existence of more efficient methods. In this work, we introduce SwiFTeR, an architecture almost entirely based on Fourier Transform and Retention, to tackle the main efficiency bottlenecks of current light image captioning models, being the visual backbone's onerosity, and the decoder's quadratic cost. SwiFTeR is made of only 20M parameters, and requires 3.1 GFLOPs for a single forward pass. Additionally, it showcases superior scalability to the caption length and its small memory requirements enable more images to be processed in parallel, compared to the traditional transformer-based architectures. For instance, it can generate 400 captions in one second. Although, for the time being, the caption quality is lower (110.2 CIDEr-D), most of the decrease is not attributed to the architecture but rather an incomplete training practice which currently leaves much room for improvements. Overall, SwiFTeR points toward a promising direction to new efficient architectural design. The implementation code will be released in the future., Comment: Pre-print version of paper accepted for ICONIP 2024

Autoregressive Sequence-To-Sequence models are the foundation of many Deep Learning achievements in major research fields such as Vision and Natural Language Processing. Despite that, they still present significant limitations. For instance, when errors occur in the early steps of the prediction, the whole output is severely affected. Such reliance on previously predicted tokens and the inherent computational unfriendliness of sequential algorithms, motivated researchers to explore different architectures and methods in the search for bidirectional approaches. In this work, we introduce the Bidirectional Awareness Induction (BAI), a training method that leverages a subset of elements in the network, the Pivots, to perform bidirectional learning without breaking the autoregressive constraints. To showcase its flexibility, we apply the method to three architectures, the Transformer, ExpansionNet v2 and GPT, then perform experiments over three tasks. Experimental results showcase BAI's effectiveness on all selected tasks and architectures. In particular, we observed an increase of up to 2.4 CIDEr in Image-Captioning, 4.96 BLEU in Neural Machine Translation, and 1.16 ROUGE in Text Summarization compared to the respective baselines. Notably, BAI not only has a positive impact on models trained from scratch but on pre-trained models as well. Such an aspect, combined with the absence of architectural requirements synergizes well with the current trend of LLMs.

Accurate ocean dynamics modeling is crucial for enhancing understanding of ocean circulation, predicting climate variability, and tackling challenges posed by climate change. Despite improvements in traditional numerical models, predicting global ocean variability over multi-year scales remains challenging. Here, we propose ORCA-DL (Oceanic Reliable foreCAst via Deep Learning), the first data-driven 3D ocean model for seasonal to decadal prediction of global ocean circulation. ORCA-DL accurately simulates three-dimensional ocean dynamics and outperforms state-of-the-art dynamical models in capturing extreme events, including El Ni\~no-Southern Oscillation and upper ocean heatwaves. This demonstrates the high potential of data-driven models for efficient and accurate global ocean forecasting. Moreover, ORCA-DL stably emulates ocean dynamics at decadal timescales, demonstrating its potential even for skillful decadal predictions and climate projections.

We introduce an interpretable-by-design method, optimized model-analog, that integrates deep learning with model-analog forecasting which generates forecasts from similar initial climate states in a repository of model simulations. This hybrid framework employs a convolutional neural network to estimate state-dependent weights to identify initial analog states that lead to shadowing target trajectories. The advantage of our method lies in its inherent interpretability, offering insights into initial-error-sensitive regions through estimated weights and the ability to trace the physically-based evolution of the system through analog forecasting. We evaluate our approach using the Community Earth System Model Version 2 Large Ensemble to forecast the El Ni\~no-Southern Oscillation (ENSO) on a seasonal-to-annual time scale. Results show a 10% improvement in forecasting equatorial Pacific sea surface temperature anomalies at 9-12 months leads compared to the unweighted model-analog technique. Furthermore, our model demonstrates improvements in boreal winter and spring initialization when evaluated against a reanalysis dataset. Our approach reveals state-dependent regional sensitivity linked to various seasonally varying physical processes, including the Pacific Meridional Modes, equatorial recharge oscillator, and stochastic wind forcing. Additionally, forecasts of El Ni\~no and La Ni\~na are sensitive to different initial states: El Ni\~no forecasts are more sensitive to initial error in tropical Pacific sea surface temperature in boreal winter, while La Ni\~na forecasts are more sensitive to initial error in tropical Pacific zonal wind stress in boreal summer. This approach has broad implications for forecasting diverse climate phenomena, including regional temperature and precipitation, which are challenging for the model-analog approach alone.

Although the Transformer is currently the best-performing architecture in the homogeneous configuration (self-attention only) in Neural Machine Translation, many State-of-the-Art models in Natural Language Processing are made of a combination of different Deep Learning approaches. However, these models often focus on combining a couple of techniques only and it is unclear why some methods are chosen over others. In this work, we investigate the effectiveness of integrating an increasing number of heterogeneous methods. Based on a simple combination strategy and performance-driven synergy criteria, we designed the Multi-Encoder Transformer, which consists of up to five diverse encoders. Results showcased that our approach can improve the quality of the translation across a variety of languages and dataset sizes and it is particularly effective in low-resource languages where we observed a maximum increase of 7.16 BLEU compared to the single-encoder model.

Electromagnetic waves possessing orbital angular momentum (OAM) are powerful tools for applications in optical communications, new quantum technologies and optical tweezers. Recently, they have attracted growing interest since they can be harnessed to detect peculiar helical dichroic effects in chiral molecular media and in magnetic nanostructures. In this work, we perform single-shot per position ptychography on a nanostructured object at a seeded free-electron laser, using extreme ultraviolet OAM beams of different topological charge order $\ell$ generated with spiral zone plates. By controlling $\ell$, we demonstrate how the structural features of OAM beam profile determine an improvement of about 30% in image resolution with respect to conventional Gaussian beam illumination. This result extends the capabilities of coherent diffraction imaging techniques, and paves the way for achieving time-resolved high-resolution (below 100 nm) microscopy on large area samples., Comment: M. Pancaldi and F. Guzzi contributed equally to this work

Memory interference may heavily inflate task execution times in Heterogeneous Systems-on-Chips (HeSoCs). Knowing worst-case interference is consequently fundamental for supporting the correct execution of time-sensitive applications. In most of the literature, worst-case interference is assumed to be generated by, and therefore is estimated through read-intensive synthetic workloads with no caching. Yet these workloads do not always generate worst-case interference. This is the consequence of the general results reported in this work. By testing on multiple architectures, we determined that the highest interference generation traffic pattern is actually hardware dependant, and that making assumptions could lead to a severe underestimation of the worst-case (in our case, of more than 9x)., Comment: Accepted for presentation at the CPS workshop 2023 (http://www.cpsschool.eu/cps-workshop)

El Ni\~no Southern Oscillation (ENSO) diversity is characterized based on the longitudinal location of maximum sea surface temperature anomalies (SSTA) and amplitude in the tropical Pacific, as Central Pacific (CP) events are typically weaker than Eastern Pacific (EP) events. SSTA pattern and intensity undergo low-frequency modulations, affecting ENSO prediction skill and remote impacts. Yet, how different ENSO types contribute to these decadal variations and long-term variance trends remain uncertain. Here, we decompose the low-frequency changes of ENSO variance into contributions from ENSO diversity categories. We propose a fuzzy clustering of monthly SSTA to allow for non-binary event category memberships. Our approach identifies two La Ni\~na and three El Ni\~no categories and shows that the shift of ENSO variance in the mid-1970s is associated with an increasing likelihood of strong La Ni\~na and extreme El Ni\~no events.

Studies of light-induced demagnetization started with the experiment performed by Beaupaire et al. on nickel. Here, we present theoretical predictions for X-ray induced demagnetization of nickel, with X-ray photon energies tuned to its $M_3$ and $L_3$ absorption edges. We show that the specific feature in the density of states of the d-band of Ni, a sharp peak located just above the Fermi level, strongly influences the change of the predicted magnetic signal, making it stronger than in the previously studied case of cobalt. We believe that this finding will inspire future experiments on magnetic processes in X-ray irradiated nickel., Comment: 6 pages, 5 figures (10 panels), 1 table, 30 references; pdfRevTeX class; double-column formatting

The Image Captioning research field is currently compromised by the lack of transparency and awareness over the End-of-Sequence token () in the Self-Critical Sequence Training. If the token is omitted, a model can boost its performance up to +4.1 CIDEr-D using trivial sentence fragments. While this phenomenon poses an obstacle to a fair evaluation and comparison of established works, people involved in new projects are given the arduous choice between lower scores and unsatisfactory descriptions due to the competitive nature of the research. This work proposes to solve the problem by spreading awareness of the issue itself. In particular, we invite future works to share a simple and informative signature with the help of a library called SacreEOS. Code available at \emph{\href{https://github.com/jchenghu/sacreeos}{https://github.com/jchenghu/sacreeos}}

Time-resolved X-ray Emission/Absorption Spectroscopy (Tr-XES/XAS) is an informative experimental tool sensitive to electronic dynamics in materials, widely exploited in diverse research fields. Typically, Tr-XES/XAS requires X-ray pulses with both a narrow bandwidth and sub-picosecond pulse duration, a combination that in principle finds its optimum with Fourier transform-limited pulses. In this work, we explore an alternative xperimental approach, capable of simultaneously retrieving information about unoccupied (XAS) and occupied (XES) states from the stochastic fluctuations of broadband extreme ultraviolet pulses of a free-electron laser. We used this method, in combination with singular value decomposition and Tikhonov regularization procedures, to determine the XAS/XES response from a crystalline silicon sample at the L2,3-edge, with an energy resolution of a few tens of meV. Finally, we combined this spectroscopic method with a pump-probe approach to measure structural and electronic dynamics of a silicon membrane. Tr-XAS/XES data obtained after photoexcitation with an optical laser pulse at 390 nm allowed us to observe perturbations of the band structure, which are compatible with the formation of the predicted precursor state of a non-thermal solid-liquid phase transition associated with a bond softening phenomenon.

The possibility to obtain a three-dimensional representation of a single object with sub-$\mu$m resolution is crucial in many fields, from material science to clinical diagnostics. This is typically achieved through tomography, which combines multiple two-dimensional images of the same object captured at different orientations. However, this serial imaging method prevents single-shot acquisition in imaging experiments at free electron lasers. In the present experiment, we report on a new approach to 3D imaging using extreme-ultraviolet radiation. In this method, two EUV pulses hit simultaneously an isolated 3D object from different sides, generating independent coherent diffraction patterns, resulting in two distinct bidimensional views obtained via phase retrieval. These views are then used to obtain a 3D reconstruction using a ray tracing algorithm. This EUV stereoscopic imaging approach, similar to the natural process of binocular vision, provides sub-$\mu$m spatial resolution and single shot capability. Moreover, ultrafast time resolution and spectroscopy can be readily implemented, a further extension to X-ray wavelengths can be envisioned as well., Comment: Authors Danny Fainozzi and Matteo Ippoliti contributed equally

Time-resolved ultrafast EUV magnetic scattering was used to test a recent prediction of >10 km/s domain wall speeds by optically exciting a magnetic sample with a nanoscale labyrinthine domain pattern. Ultrafast distortion of the diffraction pattern was observed at markedly different timescales compared to the magnetization quenching. The diffraction pattern distortion shows a threshold-dependence with laser fluence, not seen for magnetization quenching, consistent with a picture of domain wall motion with pinning sites. Supported by simulations, we show that a speed of $\approx$ 66 km/s for highly curved domain walls can explain the experimental data. While our data agree with the prediction of extreme, non-equilibrium wall speeds locally, it differs from the details of the theory, suggesting that additional mechanisms are required to fully understand these effects., Comment: 5 pages, 4 figures; Supplemental Material: 9 pages, 9 figures

Coherent light-matter interactions mediated by opto-magnetic phenomena like the inverse Faraday effect (IFE) are expected to provide a non-thermal pathway for ultrafast manipulation of magnetism on timescales as short as the excitation pulse itself. As the IFE scales with the spin-orbit coupling strength of the involved electronic states, photo-exciting the strongly spin-orbit coupled core-level electrons in magnetic materials appears as an appealing method to transiently generate large opto-magnetic moments. Here, we investigate this scenario in a ferrimagnetic GdFeCo alloy by using intense and circularly polarized pulses of extreme ultraviolet radiation. Our results reveal ultrafast and strong helicity-dependent magnetic effects which are in line with the characteristic fingerprints of an IFE, corroborated by ab initio opto-magnetic IFE theory and atomistic spin dynamics simulations.

The recently discovered topological fluctuation state provides a fascinating new perspective on the ultrafast emergence of topology in condensed matter systems. However, rather little is known about the physics of this state and the origin of the topological fluctuations. Using time-resolved small-angle x-ray scattering, we observe that topological fluctuation states appear after laser excitation even if the final state does not host stable skyrmions. Simulations support these findings and reveal that the fluctuations originate from the competition between spontaneous nucleation and decay of skyrmions, consistent with Arrhenius-like activation over a potential barrier. Stable skyrmions can freeze out of such fluctuations when the effective temperature of the system relaxes faster than the decay time of the skyrmions. Our results reveal a robust scenario for the generation of topological fluctuation states, potentially enabling their study in a wide variety of magnetic systems.

Collective lattice dynamics determine essential aspects of condensed matter, such as elastic and thermal properties. These exhibit strong dependence on the length-scale, reflecting the marked wavevector dependence of lattice excitations. The extreme ultraviolet transient grating (EUV TG) approach has demonstrated the potential of accessing a wavevector range corresponding to the 10s of nm length-scale, representing a spatial scale of the highest relevance for fundamental physics and forefront technology, previously inaccessible by optical TG and other inelastic scattering methods. In this manuscript we report on the capabilities of this technique in the context of probing thermoelastic properties of matter, both in the bulk and at the surface, as well as discussing future developments and practical considerations.

IoT applications span a wide range in performance and memory footprint, under tight cost and power constraints. High-end applications rely on power-hungry Systems-on-Chip (SoCs) featuring powerful processors, large LPDDR/DDR3/4/5 memories, and supporting full-fledged Operating Systems (OS). On the contrary, low-end applications typically rely on Ultra-Low-Power ucontrollers with a "close to metal" software environment and simple micro-kernel-based runtimes. Emerging applications and trends of IoT require the "best of both worlds": cheap and low-power SoC systems with a well-known and agile software environment based on full-fledged OS (e.g., Linux), coupled with extreme energy efficiency and parallel digital signal processing capabilities. We present HULK-V: an open-source Heterogeneous Linux-capable RISC-V-based SoC coupling a 64-bit RISC-V processor with an 8-core Programmable Multi-Core Accelerator (PMCA), delivering up to 13.8 GOps, up to 157 GOps/W and accelerating the execution of complex DSP and ML tasks by up to 112x over the host processor. HULK-V leverages a lightweight, fully digital memory hierarchy based on HyperRAM IoT DRAM that exposes up to 512 MB of DRAM memory to the host CPU. Featuring HyperRAMs, HULK-V doubles the energy efficiency without significant performance loss compared to featuring power-hungry LPDDR memories, requiring expensive and large mixed-signal PHYs. HULK-V, implemented in Global Foundries 22nm FDX technology, is a fully digital ultra-low-cost SoC running a 64-bit Linux software stack with OpenMP host-to-PMCA offload within a power envelope of just 250 mW., Comment: This paper has been accepted as full paper at DATE23 https://www.date-conference.com/date-2023-accepted-papers#Regular-Papers

The electronic excitation occurring on adsorbates at ultrafast time scales from optical lasers that initiate surface chemical reactions is still an open question. Here, we report the ultrafast temporal evolution of X-ray absorption spectroscopy (XAS) and X-ray emission spectroscopy (XES) of a simple well known adsorbate prototype system, namely carbon (C) atoms adsorbed on a nickel (Ni(100)) surface, following intense laser optical pumping at 400 nm. We observe ultrafast (~100 fs) changes in both XAS and XES showing clear signatures of the formation of a hot electron-hole pair distribution on the adsorbate. This is followed by slower changes on a few ps time scale, shown to be consistent with thermalization of the complete C/Ni system. Density functional theory spectrum simulations support this interpretation., Comment: 33 pages, 12 figures. Submitted to Physical Review Letters

We demonstrate the measurement of ultrafast dynamics using ghost spectroscopy and a pump-probe approach with an optical pump and a short-wavelength radiation probe. The ghost spectroscopy approach is used to overcome the challenge of the strong intensity and spectrum fluctuations at free-electron lasers and to provide high -spectral resolution, which enables the measurement of small energy shifts in the absorption spectrum. We exploit the high resolution to explore the dynamics of the charge carrier excitations and relaxations and their impact on the photoinduced structural changes in silicon by measuring the variation of the absorption spectrum of a Si(100) membrane near the silicon L2,3 edge and the accompanying edge shifts in response to the optical illumination.

We introduce a method called the Expansion mechanism that processes the input unconstrained by the number of elements in the sequence. By doing so, the model can learn more effectively compared to traditional attention-based approaches. To support this claim, we design a novel architecture ExpansionNet v2 that achieved strong results on the MS COCO 2014 Image Captioning challenge and the State of the Art in its respective category, with a score of 143.7 CIDErD in the offline test split, 140.8 CIDErD in the online evaluation server and 72.9 AllCIDEr on the nocaps validation set. Additionally, we introduce an End to End training algorithm up to 2.8 times faster than established alternatives. Source code available at: https://github.com/jchenghu/ExpansionNet_v2

Magnetization dynamics triggered with ultrashort laser pulses has been attracting significant attention, with strong focus on the dynamics excited by VIS/NIR pulses. Only recently, strong magnetic response in solid materials induced by intense X-ray pulses from free-electron lasers (FELs) has been observed. The exact mechanisms that trigger the X-ray induced demagnetization are not yet fully understood. They are subject of on-going experimental and theoretical investigations. Here, we present a theoretical analysis of electronic processes occurring during demagnetization of Co multilayer system irradiated by X-ray pulses tuned to L$_3$-absorption edge of cobalt. We show that, similarly as in the case of X-ray induced demagnetization at M-edge of Co, electronic processes play a predominant role in the demagnetization until the pulse fluence does not exceed the structural damage threshold. The impact of electronic processes can reasonably well explain the available experimental data, without a need to introduce the mechanism of stimulated elastic forward scattering., Comment: 10 pages, 4 figures (7 panels), 57 references; pdfRevTeX class; double column formatting; two appendices and 18 references added; title changed; author-created version submitted to and accepted/published in Physical Review B journal. The text of the appendices (including detailed description of the model used) is similar to some parts of arXiv:2202.13845

Most recent state of the art architectures rely on combinations and variations of three approaches: convolutional, recurrent and self-attentive methods. Our work attempts in laying the basis for a new research direction for sequence modeling based upon the idea of modifying the sequence length. In order to do that, we propose a new method called "Expansion Mechanism" which transforms either dynamically or statically the input sequence into a new one featuring a different sequence length. Furthermore, we introduce a novel architecture that exploits such method and achieves competitive performances on the MS-COCO 2014 data set, yielding 134.6 and 131.4 CIDEr-D on the Karpathy test split in the ensemble and single model configuration respectively and 130 CIDEr-D in the official online evaluation server, despite being neither recurrent nor fully attentive. At the same time we address the efficiency aspect in our design and introduce a convenient training strategy suitable for most computational resources in contrast to the standard one. Source code is available at https://github.com/jchenghu/exploring

Ultrafast control of magnetization on the nanometer length scale, in particular all-optical switching, is key to putting ultrafast magnetism on the path towards future technological application in data storage technology. However, magnetization manipulation with light on this length scale is challenging due to the wavelength limitations of optical radiation. Here, we excite transient magnetic gratings in a GdFe alloy with a periodicity of 87 nm by interference of two coherent femtosecond light pulses in the extreme ultraviolet spectral range at the free electron laser facility FERMI. The subsequent ultrafast evolution of the magnetization pattern is probed by diffraction of a third, time-delayed pulse tuned to the Gd N-edge at a wavelength of 8.3 nm. By examining the simultaneously recorded first and second diffraction orders and by performing reference real-space measurements with a wide-field magneto-optical microscope with femtosecond time resolution, we can conclusively demonstrate the ultrafast emergence of all-optical switching on the nanometer length scale., Comment: 6 pages, 5 figures