Search

Matrix-vector multiplications (MVMs) are essential for a wide range of applications, particularly in modern machine learning and quantum computing. In photonics, there is growing interest in developing architectures capable of performing linear operations with high speed, low latency, and minimal loss. Traditional interferometric photonic architectures, such as the Clements design, have been extensively used for MVM operations. However, as these architectures scale, improving stability and robustness becomes critical. In this paper, we introduce a novel photonic braid interferometer architecture that outperforms both the Clements and Fldzhyan designs in these aspects. Using numerical simulations, we evaluate the performance of these architectures under ideal conditions and systematically introduce non-idealities such as insertion losses, beam splitter imbalances, and crosstalk. The results demonstrate that the braid architecture offers superior robustness due to its symmetrical design and reduced layer count. Further analysis shows that the braid architecture is particularly advantageous in large-scale implementations, delivering better performance as the size of the interferometer increases. We also assess the footprint and total insertion losses of each architecture. Although waveguide crossings in the braid architecture slightly increase the footprint and insertion loss, recent advances in crossing technology significantly minimize these effects. Our study suggests that the braid architecture is a robust solution for photonic neuromorphic computing, maintaining high fidelity in realistic conditions where imperfections are inevitable.

This paper describes speech enhancement for realtime automatic speech recognition (ASR) in real environments. A standard approach to this task is to use neural beamforming that can work efficiently in an online manner. It estimates the masks of clean dry speech from a noisy echoic mixture spectrogram with a deep neural network (DNN) and then computes a enhancement filter used for beamforming. The performance of such a supervised approach, however, is drastically degraded under mismatched conditions. This calls for run-time adaptation of the DNN. Although the ground-truth speech spectrogram required for adaptation is not available at run time, blind dereverberation and separation methods such as weighted prediction error (WPE) and fast multichannel nonnegative matrix factorization (FastMNMF) can be used for generating pseudo groundtruth data from a mixture. Based on this idea, a prior work proposed a dual-process system based on a cascade of WPE and minimum variance distortionless response (MVDR) beamforming asynchronously fine-tuned by block-online FastMNMF. To integrate the dereverberation capability into neural beamforming and make it fine-tunable at run time, we propose to use weighted power minimization distortionless response (WPD) beamforming, a unified version of WPE and minimum power distortionless response (MPDR), whose joint dereverberation and denoising filter is estimated using a DNN. We evaluated the impact of run-time adaptation under various conditions with different numbers of speakers, reverberation times, and signal-to-noise ratios (SNRs)., Comment: Accepted to APSIPA2024

Security and RAS are two non-functional requirements under focus for current systems developed for the computing continuum. Due to the increased number of interconnected computer systems across the continuum, security becomes especially pervasive at all levels, from the smallest edge device to the high-performance cloud at the other end. Similarly, RAS (Reliability, Availability, and Serviceability) ensures the robustness of a system towards hardware defects. Namely, making them reliable, with high availability and design for easy service. In this paper and as a result of the Vitamin-V EU project, the authors detail the comprehensive approach to malware and hardware attack detection; as well as, the RAS features envisioned for future systems across the computing continuum.

In the United States, K-12 school finance is largely controlled by the states. Every year, hundreds of billions of dollars in public funds are distributed based on 51 different configurations of formulas, rules, and regulations to over 13,000 districts that vary in terms of the students they serve, their ability to raise revenue locally, and many other characteristics. Good school finance systems compensate for factors states cannot control (e.g., student poverty, labor costs) using levers that they can control (e.g., driving funding to districts that need it most). The authors have devised a framework that evaluates states based largely on how well they accomplish this balance. Each state's funding is assessed while accounting directly for the students and communities served by its public schools. In this sixth edition of the annual report, the K-12 school finance systems of all 50 states and the District of Columbia is evaluated. The latest year of data presented is the 2020-21 school year. [This report was co-prepared by University of Miami School of Education and Human Development.]

Merger rate predictions of Massive Black Hole (MBH) seeds from large-scale cosmological simulations differ widely, with recent studies highlighting the challenge of low-mass MBH seeds failing to reach the galactic center, a phenomenon known as the seed sinking problem. In this work, we tackle this issue by integrating cosmological simulations and galaxy merger simulations from the MAGICS-I and MAGICS-II resimulation suites with high-resolution $N$-body simulations. Building on the findings of MAGICS-II, which showed that only MBH seeds embedded in stellar systems are able to sink to the center, we extend the investigation by incorporating nuclear star clusters (NSCs) into our models. Utilizing $N$-body resimulations with up to $10^7$ particles, we demonstrate that interactions between NSCs and their surrounding galactic environment, particularly tidal forces triggered by cluster interactions, significantly accelerate the sinking of MBHs to the galactic center. This process leads to the formation of a hard binary in $\lesssim 500$ Myr after the onset of a galaxy merger. Our results show that in 8 out of 12 models, the high stellar density of the surrounding NSCs enhances MBH hardening, facilitating gravitational wave (GW) mergers by redshift $z = 4$. We conclude that at $z > 4$, dense NSCs serve as the dominant channel for MBH seed mergers, producing a merger rate of $0.3$--$0.6\, \mathrm{yr}^{-1}$ at $z = 4$, which is approximately 300--600 times higher than in non-NSC environments. In contrast, in environments without NSCs, surrounding dark matter plays a more significant role in loss-cone scattering., Comment: 26 pages, 13 figures. Submitted to ApJ. Comments welcome!

Context. Computational astronomy has reached the stage where running a gravitational N-body simulation of a stellar system, such as a Milky Way star cluster, is computationally feasible, but a major limiting factor that remains is the ability to set up physically realistic initial conditions. Aims. We aim to obtain realistic initial conditions for N-body simulations by taking advantage of machine learning, with emphasis on reproducing small-scale interstellar distance distributions. Methods. The computational bottleneck for obtaining such distance distributions is the hydrodynamics of star formation, which ultimately determine the features of the stars, including positions, velocities, and masses. To mitigate this issue, we introduce a new method for sampling physically realistic initial conditions from a limited set of simulations using Gaussian processes. Results. We evaluated the resulting sets of initial conditions based on whether they meet tests for physical realism. We find that direct sampling based on the learned distribution of the star features fails to reproduce binary systems. Consequently, we show that physics-informed sampling algorithms solve this issue, as they are capable of generating realisations closer to reality., Comment: Accepted to A&A (4th Sep 2024)

We do experiments on physical agents with dynamic binary ideologies, deep memories of previous probes of neigboring agents, but fixed personalities that interpret the memory content to make ideological decisions. We find experimentally a critical memory depth below which complete ideological polarization of the collective cannot occur, and above which it is inevitable, an emergent symmetry breaking that is memory depth dependent. Depending on the details of the personalities, the polarization can be static or dynamic in time, even in certain cases chaotic due to nonreciprocity in how the agents respond to other agents. Thus, agents with different personalities and depths of memory serve as a physics analog of the ideology dynamics among biased individuals, illuminating how decisions influenced by individual memories of past interactions can shape and influence subsequent polarization. Perhaps such applications of physics-based systems to political systems will help us to understand the ideological instabilities observed today.

Detecting monotonic changes in heart rate (HR) is crucial for early identification of cardiac conditions and health management. This is particularly important for dementia patients, where HR trends can signal stress or agitation. Developing wearable technologies that can perform always-on monitoring of HRs is essential to effectively detect slow changes over extended periods of time. However, designing compact electronic circuits that can monitor and process bio-signals continuously, and that can operate in a low-power regime to ensure long-lasting performance, is still an open challenge. Neuromorphic technology offers an energy-efficient solution for real-time health monitoring. We propose a neuromorphic implementation of a Neural State Machine (NSM) network to encode different health states and switch between them based on the input stimuli. Our focus is on detecting monotonic state switches in electrocardiogram data to identify progressive HR increases. This innovative approach promises significant advancements in continuous health monitoring and management.

Pair-instability supernovae (PISNe) have crucial implications for many astrophysical topics, including the search for very massive stars, the black hole mass spectrum, and galaxy chemical enrichment. To this end, we need to understand where PISNe are across cosmic time, and what are their favourable galactic environments. We present a new determination of the PISN rate as a function of redshift, obtained by combining up-to-date stellar evolution tracks from the PARSEC and FRANEC codes, with an up-to-date semi-empirical determination of the star formation rate and metallicity evolution of star-forming galaxies throughout cosmic history. We find the PISN rate to exhibit a huge dependence on the model assumptions, including the criterion to identify stars unstable to pair production, and the upper limit of the stellar initial mass function. Remarkably, the interplay between the maximum metallicity at which stars explode as PISNe, and the dispersion of the galaxy metallicity distribution, dominates the uncertainties, causing a $\sim$ seven-orders-of-magnitude PISN rate range. Furthermore, we show a comparison with the core-collapse supernova rate, and study the properties of the favourable PISN host galaxies. According to our results, the main contribution to the PISN rate comes from metallicities between $\sim 10^{-3}$ and $10^{-2}$, against the common assumption that views very-low-metallicity, Population III stars as exclusive or dominant PISN progenitors. The strong dependencies we find offer the opportunity to constrain stellar and galaxy evolution models based on possible future (or the lack of) PISN observations., Comment: Accepted for publication on MNRAS, 23 pages, 25 figures, 4 tables

We investigate the critical behavior of the Kinetic Ising model with non-reciprocal nearest neighbors interactions. A finite-size scaling study suggests that the model belongs to the Ising universality class. We characterize the off-equilibrium behavior of the model by measuring the entropy production rate and by studying the coarsening dynamics, which shows a super-diffusive relaxation. A coarse-grained equation of motion is in compelling agreement with the numerical findings. A one-loop renormalization group calculation shows that the model belongs to the Model A universality class.

New experimental methods make it possible to measure the expression levels of many genes, simultaneously, in snapshots from thousands or even millions of individual cells. Current approaches to analyze these experiments involve clustering or low-dimensional projections. Here we use the principle of maximum entropy to obtain a probabilistic description that captures the observed presence or absence of mRNAs from hundreds of genes in cells from the mammalian brain. We construct the Ising model compatible with experimental means and pairwise correlations, and validate it by showing that it gives good predictions for higher-order statistics. We notice that the probability distribution of cell states has many local maxima. By labeling cell states according to the associated maximum, we obtain a cell classification that agrees well with previous results that use traditional clustering techniques. Our results provide quantitative descriptions of gene expression statistics and interpretable criteria for defining cell classes, supporting the hypothesis that cell classes emerge from the collective interaction of gene expression levels., Comment: 14 pages, 21 figures

The increasing adoption of connectivity and electronic components in vehicles makes these systems valuable targets for attackers. While automotive vendors prioritize safety, there remains a critical need for comprehensive assessment and analysis of cyber risks. In this context, this paper proposes a Social Media Automotive Threat Intelligence (SOCMATI) framework, specifically designed for the emerging field of automotive cybersecurity. The framework leverages advanced intelligence techniques and machine learning models to extract valuable insights from social media. Four use cases illustrate the framework's potential by demonstrating how it can significantly enhance threat assessment procedures within the automotive industry., Comment: Position paper, four pages, two images

The automotive industry has evolved significantly since the introduction of the Ford Model T in 1908. Today's vehicles are not merely mechanical constructs; they are integral components of a complex digital ecosystem, equipped with advanced connectivity features powered by Artificial Intelligence and cloud computing technologies. This evolution has enhanced vehicle safety, efficiency, and the overall driving experience. However, it also introduces new challenges, notably in cybersecurity. With the increasing integration of digital technologies, vehicles have become more susceptible to cyber-attacks, prompting significant cybersecurity concerns. These concerns include securing sensitive data, protecting vehicles from unauthorized access, and ensuring user privacy. In response, the automotive industry is compelled to adopt robust cybersecurity measures to safeguard both vehicles and data against potential threats. Legislative frameworks such as UNR155 and UNR156 by the United Nations, along with other international regulations, aim to establish stringent cybersecurity mandates. These regulations require compliance with comprehensive cybersecurity management systems and necessitate regular updates and testing to cope with the evolving nature of cyber threats. The introduction of such regulations highlights the growing recognition of cybersecurity as a critical component of automotive safety and functionality. The future of automotive cybersecurity lies in the continuous development of advanced protective measures and collaborative efforts among all stakeholders, including manufacturers, policymakers, and cybersecurity professionals. Only through such concerted efforts can the industry hope to address the dual goals of innovation in vehicle functionality and stringent security measures against the backdrop of an increasingly interconnected digital landscape.

Among the key contributions of Vitamin-V (2023-2025 Horizon Europe project), we develop a complete RISC-V open-source software stack for cloud services with comparable performance to the cloud-dominant x86 counterpart. In this paper, we detail the software suites and applications ported plus the three cloud setups under evaluation., Comment: RISC-V Summit Europe 2024, 24-28 June 2024

This work evaluates how well hardware-based approaches detect stack buffer overflow (SBO) attacks in RISC-V systems. We conducted simulations on the PULP platform and examined micro-architecture events using semi-supervised anomaly detection techniques. The findings showed the challenge of detection performance. Thus, a potential solution combines software and hardware-based detectors concurrently, with hardware as the primary defense. The hardware-based approaches present compelling benefits that could enhance RISC-V-based architectures., Comment: 2 pages, 2 figures

Modern vehicles are increasingly vulnerable to attacks that exploit network infrastructures, particularly the Controller Area Network (CAN) networks. To effectively counter such threats using contemporary tools like Intrusion Detection Systems (IDSs) based on data analysis and classification, large datasets of CAN messages become imperative. This paper delves into the feasibility of generating synthetic datasets by harnessing the modeling capabilities of simulation frameworks such as Simulink coupled with a robust representation of attack models to present CARACAS, a vehicular model, including component control via CAN messages and attack injection capabilities. CARACAS showcases the efficacy of this methodology, including a Battery Electric Vehicle (BEV) model, and focuses on attacks targeting torque control in two distinct scenarios., Comment: 6 pages, 8 figures, TrustAICyberSec workshop - IEEE ISCC 2024

In the effort to learn from extensive collections of distributed data, federated learning has emerged as a promising approach for preserving privacy by using a gradient-sharing mechanism instead of exchanging raw data. However, recent studies show that private training data can be leaked through many gradient attacks. While previous analytical-based attacks have successfully reconstructed input data from fully connected layers, their effectiveness diminishes when applied to convolutional layers. This paper introduces an advanced data leakage method to efficiently exploit convolutional layers' gradients. We present a surprising finding: even with non-fully invertible activation functions, such as ReLU, we can analytically reconstruct training samples from the gradients. To the best of our knowledge, this is the first analytical approach that successfully reconstructs convolutional layer inputs directly from the gradients, bypassing the need to reconstruct layers' outputs. Prior research has mainly concentrated on the weight constraints of convolution layers, overlooking the significance of gradient constraints. Our findings demonstrate that existing analytical methods used to estimate the risk of gradient attacks lack accuracy. In some layers, attacks can be launched with less than 5% of the reported constraints.

Rehearsal-based Continual Learning (CL) has been intensely investigated in Deep Neural Networks (DNNs). However, its application in Spiking Neural Networks (SNNs) has not been explored in depth. In this paper we introduce the first memory-efficient implementation of Latent Replay (LR)-based CL for SNNs, designed to seamlessly integrate with resource-constrained devices. LRs combine new samples with latent representations of previously learned data, to mitigate forgetting. Experiments on the Heidelberg SHD dataset with Sample and Class-Incremental tasks reach a Top-1 accuracy of 92.5% and 92%, respectively, without forgetting the previously learned information. Furthermore, we minimize the LRs' requirements by applying a time-domain compression, reducing by two orders of magnitude their memory requirement, with respect to a naive rehearsal setup, with a maximum accuracy drop of 4%. On a Multi-Class-Incremental task, our SNN learns 10 new classes from an initial set of 10, reaching a Top-1 accuracy of 78.4% on the full test set.

We propose to investigate the time modulation of radioisotope decays deep underground as a method to explore axion dark matter. In this work, we focus on the $\alpha$-decay of heavy isotopes and develop a theoretical description for the $\theta$-dependence of $\alpha$-decay half-lives, which enables us to predict the time variation of $\alpha$-radioactivity in response to an oscillating axion dark matter background. To probe this scenario, we have recently constructed and installed a setup deep underground at the Gran Sasso Laboratory, based on the $\alpha$-decay of Americium-241. This prototype experiment, named RadioAxion-$\alpha$, will allow us to explore a broad range of oscillation's periods, from a micro-second up to one year, thus providing competitive limits on the axion decay constant across 13 orders of magnitude in the axion mass, ranging from $10^{-9}$ eV to $10^{-20}$ eV after one month of data collection, and down to $10^{-22}$ eV after three years., Comment: 7 pages, 3 figures. Added an Appendix providing an analytical treatment of the theta dependence of alpha decay. Version to appear in Physics Letters B

Gaia BH3 is the first observed dormant black hole (BH) with a mass of $\approx{30}$ M$_\odot$ and represents the first confirmation that such massive BHs are associated with metal-poor stars. Here, we explore the isolated binary formation channel for Gaia BH3 focusing on the old and metal-poor stellar population of the Milky Way halo. We use the MIST stellar models and our open-source population synthesis code SEVN to evolve $5.6 \times 10^8$ binaries exploring 20 sets of parameters. We find that systems like Gaia BH3 form preferentially from binaries initially composed of a massive star ($40-60$ M$_\odot$) and a low mass companion ($<1$ M$_\odot$) in a wide ($P>10^3$ days) and eccentric orbit ($e>0.6$). Such progenitor binary stars do not undergo any Roche-lobe overflow episode during their entire evolution, so that the final orbital properties of the BH-star system are determined at the core collapse of the primary star. Low natal kicks ($\lesssim$ 10~km/s) significantly favour the formation of Gaia BH3-like systems, but high velocity kicks up to $\approx 220$ km/s are also allowed. We estimate the formation efficiency for Gaia BH3-like systems in old ($t>10$ Gyr) and metal-poor ($Z<0.01$) populations to be $\sim 4 \times 10^{-8}$ M$_\odot^{-1}$ (for our fiducial model), representing $\sim 3\%$ of the whole simulated BH-star population. We expect up to $\approx 4000$ BH-star systems in the Galactic halo formed through isolated evolution, of which $\approx 100$ are compatible with Gaia BH3-like. Given the density profile of the Galactic halo we do not expect more than one at the observed distance of Gaia BH3. Our models show that, even if it was born inside a stellar cluster, Gaia BH3 is compatible with a primordial binary star that escaped from its parent cluster without experiencing significant dynamical interactions., Comment: Accepted for publication in A&A. The produced datasets are publicly available in Zenodo. I Initiail conditions and run scripts: https://zenodo.org/records/11617742; IIA simulation outcomes: https://zenodo.org/records/11992265; IIB simulation outcomes, different metallicities: https://zenodo.org/records/12188369; IIC simulation outcomes, fiducial 1E8, https://zenodo.org/records/12189734

One of today's main concerns is to bring Artificial Intelligence power to embedded systems for edge applications. The hardware resources and power consumption required by state-of-the-art models are incompatible with the constrained environments observed in edge systems, such as IoT nodes and wearable devices. Spiking Neural Networks (SNNs) can represent a solution in this sense: inspired by neuroscience, they reach unparalleled power and resource efficiency when run on dedicated hardware accelerators. However, when designing such accelerators, the amount of choices that can be taken is huge. This paper presents SpikExplorer, a modular and flexible Python tool for hardware-oriented Automatic Design Space Exploration to automate the configuration of FPGA accelerators for SNNs. Using Bayesian optimizations, SpikerExplorer enables hardware-centric multi-objective optimization, supporting factors such as accuracy, area, latency, power, and various combinations during the exploration process. The tool searches the optimal network architecture, neuron model, and internal and training parameters, trying to reach the desired constraints imposed by the user. It allows for a straightforward network configuration, providing the full set of explored points for the user to pick the trade-off that best fits the needs. The potential of SpikExplorer is showcased using three benchmark datasets. It reaches 95.8% accuracy on the MNIST dataset, with a power consumption of 180mW/image and a latency of 0.12 ms/image, making it a powerful tool for automatically optimizing SNNs.

School finance debates frequently turn on two crucial questions: (1) How much do state and local governments spend on K-12 education?; and (2) How are education dollars distributed across jurisdictions? This focus makes sense because the answers to these questions determine how well states are able to provide an adequate, equal education for all students. This report, however, focuses on two different questions: (1) Where do school revenues come from?; and (2) How does revenue composition affect education policymaking? Put differently, this report focuses on the degree to which states rely on state versus local revenue, or on different types of taxes (e.g., those on income, sales, or property), and how such features matter for the adequacy and equity of education spending. Two states with identical finance systems and student populations, and which spend the same amount overall on K-12 schools, might have very different adequacy and equity outcomes depending on the composition of their revenue. In short, money matters; but where money comes from matters too.

As artificial neural networks become increasingly integrated into safety-critical systems such as autonomous vehicles, devices for medical diagnosis, and industrial automation, ensuring their reliability in the face of random hardware faults becomes paramount. This paper introduces SpikingJET, a novel fault injector designed specifically for fully connected and convolutional Spiking Neural Networks (SNNs). Our work underscores the critical need to evaluate the resilience of SNNs to hardware faults, considering their growing prominence in real-world applications. SpikingJET provides a comprehensive platform for assessing the resilience of SNNs by inducing errors and injecting faults into critical components such as synaptic weights, neuron model parameters, internal states, and activation functions. This paper demonstrates the effectiveness of Spiking-JET through extensive software-level experiments on various SNN architectures, revealing insights into their vulnerability and resilience to hardware faults. Moreover, highlighting the importance of fault resilience in SNNs contributes to the ongoing effort to enhance the reliability and safety of Neural Network (NN)-powered systems in diverse domains.

Efficient semi-transparent solar cells can trigger the adoption of building integrated photovoltaics. Halide perovskites are particularly suitable in this respect owing to their tunable bandgap. Main drawbacks in the development of transparent perovskite solar cells are the high Voc deficit and the difficulties in depositing thin films over large area substrates, given the low solubility of bromide and chloride precursors. In this work, we develop a 2D and passivation strategies for the high band-gap Br perovskite able to reduce charge recombination and consequently improving the open-circuit voltage. We demonstrate 1cm 2 perovskite solar cells with Voc up to 1.73 V (1.83 eV QFLS) and a PCE of 8.2%. The AVT exceeds 70% by means of a bifacial light management and a record light utilization efficiency of 5.72 is achieved, setting a new standard for transparent photovoltaics. Moreover, we show the high ceiling of our technology towards IoT application due to a bifaciality factor of 87% along with 17% PCE under indoor lighting. Finally, the up-scaling has been demonstrated fabricating 20cm 2 -active area modules with PCE of 7.3% and Voc per cell up to 1.65V.

In contemporary times, the increasing complexity of the system poses significant challenges to the reliability, trustworthiness, and security of the SACRES. Key issues include the susceptibility to phenomena such as instantaneous voltage spikes, electromagnetic interference, neutron strikes, and out-of-range temperatures. These factors can induce switch state changes in transistors, resulting in bit-flipping, soft errors, and transient corruption of stored data in memory. The occurrence of soft errors, in turn, may lead to system faults that can propel the system into a hazardous state. Particularly in critical sectors like automotive, avionics, or aerospace, such malfunctions can have real-world implications, potentially causing harm to individuals. This paper introduces a novel fault injector designed to facilitate the monitoring, aggregation, and examination of micro-architectural events. This is achieved by harnessing the microprocessor's PMU and the debugging interface, specifically focusing on ensuring the repeatability of fault injections. The fault injection methodology targets bit-flipping within the memory system, affecting CPU registers and RAM. The outcomes of these fault injections enable a thorough analysis of the impact of soft errors and establish a robust correlation between the identified faults and the essential timing predictability demanded by SACRES.

Renormalization group theory is a powerful and intriguing technique with a wide range of applications. One of the main successes of renormalization group theory is the description of continuous phase transitions and the development of scaling theory. Most courses on phase transitions focus on scaling and critical exponents, while less attention is paid to universality, renormalization group flow, and the existence of a unique fixed point, which are the ultimate reasons why scaling theory is so effective in describing continuous phase transitions. We use a combination of Monte Carlo simulations and real space renormalization group theory to determine the renormalization group flow and to show the existence of a universal fixed point in the context of the ferromagnetic Ising model.

Including Artificial Neural Networks in embedded systems at the edge allows applications to exploit Artificial Intelligence capabilities directly within devices operating at the network periphery. This paper introduces Spiker+, a comprehensive framework for generating efficient, low-power, and low-area customized Spiking Neural Networks (SNN) accelerators on FPGA for inference at the edge. Spiker+ presents a configurable multi-layer hardware SNN, a library of highly efficient neuron architectures, and a design framework, enabling the development of complex neural network accelerators with few lines of Python code. Spiker+ is tested on two benchmark datasets, the MNIST and the Spiking Heidelberg Digits (SHD). On the MNIST, it demonstrates competitive performance compared to state-of-the-art SNN accelerators. It outperforms them in terms of resource allocation, with a requirement of 7,612 logic cells and 18 Block RAMs (BRAMs), which makes it fit in very small FPGA, and power consumption, draining only 180mW for a complete inference on an input image. The latency is comparable to the ones observed in the state-of-the-art, with 780us/img. To the authors' knowledge, Spiker+ is the first SNN accelerator tested on the SHD. In this case, the accelerator requires 18,268 logic cells and 51 BRAM, with an overall power consumption of 430mW and a latency of 54 us for a complete inference on input data. This underscores the significance of Spiker+ in the hardware-accelerated SNN landscape, making it an excellent solution to deploy configurable and tunable SNN architectures in resource and power-constrained edge applications.

Approximate Computing (AxC) techniques have become increasingly popular in trading off accuracy for performance gains in various applications. Selecting the best AxC techniques for a given application is challenging. Among proposed approaches for exploring the design space, Machine Learning approaches such as Reinforcement Learning (RL) show promising results. In this paper, we proposed an RL-based multi-objective Design Space Exploration strategy to find the approximate versions of the application that balance accuracy degradation and power and computation time reduction. Our experimental results show a good trade-off between accuracy degradation and decreased power and computation time for some benchmarks.

In the contemporary security landscape, the incorporation of photonics has emerged as a transformative force, unlocking a spectrum of possibilities to enhance the resilience and effectiveness of security primitives. This integration represents more than a mere technological augmentation; it signifies a paradigm shift towards innovative approaches capable of delivering security primitives with key properties for low-power systems. This not only augments the robustness of security frameworks, but also paves the way for novel strategies that adapt to the evolving challenges of the digital age. This paper discusses the security layers and related services that will be developed, modeled, and evaluated within the Horizon Europe NEUROPULS project. These layers will exploit novel implementations for security primitives based on physical unclonable functions (PUFs) using integrated photonics technology. Their objective is to provide a series of services to support the secure operation of a neuromorphic photonic accelerator for edge computing applications., Comment: 6 pages, 4 figures

Metal halide perovskite solar cells being one of the fastest emerging technologies for renewable energy still has to become more industry friendly in a way that will allow using it for thin film modules or tandems with conventional silicon devices. The simplest way to achieve this is to use chemical vapor deposition (CVD) technique for tandem production. In this work, we show a method for a single step production of MAPbI3 films in a simple two-zone CVD reactor from lead diacetate and methyl-ammonium iodide powders. Obtained films show highly ordered cubic MAPbI3 phase with a thickness of 400-500 um, good photoluminescence response and absorption band edge similar to spin-coated film. We used those films to produce p-i-n solar cells with ITO/NiO/MAPbI3/C60/Cu structure. The best cell showed negligible 0.23 % PCE right after the manufacturing but significantly improved to 5.5 % PCE after 8 hours of storage in the dark. The main limiting factor affecting the efficiency is low current density, which we attribute to non-optimized growth conditions; however, our approach is a first step to a single step CVD deposition of MAPbI3 on any type of substrates including texturized silicon subcells for better overall efficiency.

A novel triphenylamine-based hole transport material (HTM) with a carboxyl anchoring group (TPATC) was developed for tuning the interface between nanocrystalline NiO and double cation CsCH3(NH2)2PbI3-xClx absorber in p-i-n device architectures. We present a unique comprehensive confinement study of PSCs with TPATC as the self-assembled HTM, including analysis of numerical defect parameters, phase composition evolution, and up-scaling capabilities. Our investigation shows that the ultrathin TPATC interlayer effectively passivates traps, increases work-function of HTL by about ~0.2 eV, and enhances the charge carrier extraction efficiency. Advanced transient spectroscopy measurements revealed that modification of the NiO surface with TPATC in perovskite solar cells (PSCs) reduces the concentration of ionic defects by an order of magnitude. Interface engineering with TPATC allowed to reach power conversion efficiency of 20.58% for small area devices (0.15 cm2) under standard AM 1.5 G conditions. Using TPATC interlayer also provided stabilized performance of PSCs under operation conditions and improved sustainability of the perovskite absorber to decomposition. After continuous light-soaking (1000 h, ISOS-L-2 protocol), NiO/TPATC devices showed a slight decrease of 2% in maximum power. In contrast, NiO PSCs demonstrated decrease in power output (>20%) after 400 h. We explored the potential of TPATC to modify interfaces in large-area perovskite solar modules (PSM, active area-64.8 cm2, 12 sub-cells). By applying slot-die-coated TPATC, the PCE at AM 1.5 G conditions increased from 13.22% for NiO PSM to 15.64% for NiO/TPATC ones. This study provides new insights into the interface engineering for p-i-n perovskite solar cells, behavior of the ionic defects and their contribution to the long-term stability.

Interface engineering is one of the promising strategies for the long-term stabilization of perovskite solar cells, preventing chemical decomposition induced by external agents and promoting fast charge transfer. Recently, MXenes-2D structured transition metal carbides and nitrides with various functionalization (=O,-F,-OH) demonstrated high potential for mastering the work function in halide perovskite absorbers and significantly improved the n-type charge collection in solar cells. This work demonstrates that MXenes allow for efficient stabilization of perovskite solar cells besides improving their performances. We introduce a new mixed composite bathocuproine:MXene, i.e., (BCP:MXene) interlayer at the interface between an electron-transport layer (ETL) and a metal cathode in the p-in device structure. Our investigation demonstrates that the use of BCP:MXene interlayer slightly increases the power conversation efficiency (PCE) for PSCs (from 16.5 for reference to 17.5%) but dramatically improves the out of Glove-Box stability. Under ISOS-L-2 light soaking stress at 63$\pm$ 1.5{\textdegree}C, The T80 (time needed to reduce efficiency down to 80% of the initial one) period increased from 460 h to > 2300 h.

Kaon physics is at a turning point -- while the rare-kaon experiments NA62 and KOTO are in full swing, the end of their lifetime is approaching and the future experimental landscape needs to be defined. With HIKE, KOTO-II and LHCb-Phase-II on the table and under scrutiny, it is a very good moment in time to take stock and contemplate about the opportunities these experiments and theoretical developments provide for particle physics in the coming decade and beyond. This paper provides a compact summary of talks and discussions from the Kaons@CERN 2023 workshop., Comment: 55 pages, Summary of Kaons@CERN 23 workshop, references updated, typos fixed, version as published in EPJC