Search

Titanium (Ti) is an adhesion and contact metal commonly used in nanoelectronics and two-dimensional (2D) materials research. However, when Ti is deposited on graphene (Gr), we obtain dramatically different film morphology depending on the experimental conditions. Through a combination of transmission electron microscopy, Raman spectroscopy, and ab initio density functional theory calculations, we show that the most critical parameters are the number of Gr layers, the nature of the Gr support, and the deposition temperature. Particularly distinctive is the island morphology and large defect density of Ti on monolayer Gr, compared to bilayer or thicker Gr. We propose that this results from structural and mechanical differences between monolayer and thicker Gr flakes, where monolayer Gr is more flexible, exhibits larger surface roughness and therefore lower Ti diffusivity, and is more easily damaged. Our results highlight the extreme sensitivity of Ti morphology on Gr to processing and substrate conditions, allowing us to propose design rules for controlling Ti-Gr interface properties and morphology and to discuss the implications for other technologically relevant metal deposition processes., Comment: 11 pages, 5 figures

In missions constrained by finite resources, efficient data collection is critical. Informative path planning, driven by automated decision-making, optimizes exploration by reducing the costs associated with accurate characterization of a target in an environment. Previous implementations of active learning did not consider the action cost for regression problems or only considered the action cost for classification problems. This paper analyzes an AL algorithm for Gaussian Process regression while incorporating action cost. The algorithm's performance is compared on various regression problems to include terrain mapping on diverse simulated surfaces along metrics of root mean square error, samples and distance until convergence, and model variance upon convergence. The cost-dependent acquisition policy doesn't organically optimize information gain over distance. Instead, the traditional uncertainty metric with a distance constraint best minimizes root-mean-square error over trajectory distance. This studys impact is to provide insight into incorporating action cost with AL methods to optimize exploration under realistic mission constraints.

Hindsight experience replay (HER) is well-known to accelerate goal-based reinforcement learning (RL). While HER is generally applied to off-policy RL algorithms, we previously showed that HER can also accelerate on-policy algorithms, such as proximal policy optimization (PPO), for goal-based Predator-Prey environments. Here, we show that we can improve the previous PPO-HER algorithm by selectively applying HER in a principled manner., Comment: 11 pages, 11 Figures

Hindsight experience replay (HER) accelerates off-policy reinforcement learning algorithms for environments that emit sparse rewards by modifying the goal of the episode post-hoc to be some state achieved during the episode. Because post-hoc modification of the observed goal violates the assumptions of on-policy algorithms, HER is not typically applied to on-policy algorithms. Here, we show that HER can dramatically accelerate proximal policy optimization (PPO), an on-policy reinforcement learning algorithm, when tested on a custom predator-prey environment., Comment: 12 pages. 10 Figures

Activity-based models in transport are crucial for providing a comprehensive and realistic understanding of individuals' activity-travel patterns. Traditionally, travel surveys have been used to develop these models, but they are often costly and have small sample sizes. Mobile phone application data, one example of emerging data sources, offers an alternative with wider population coverage over extended periods for developing activity-based models. However, the challenges of using these data include sampling biases in the population coverage and individual-level data sparsity due to intermittent and irregular data collection. To synthesise activity-travel plans, we propose a novel model that combines mobile phone application data with travel survey data, addressing their limitations. Our generative model simulates multiple average weekday activity schedules for over 263,000 individuals living in Sweden, approximately 2.6% of Sweden's population. We also introduce a temporal-score approach to improve home and work location identification approaches. We assess the model's performance against an existing large-scale agent-based model of Sweden (SySMo) and a dummy model using only mobile application data. The generated activity-travel plans are comparable to the SySMo model's output and significantly surpass the dummy model's results, suggesting the proposed model's capability to generate reasonable activity-travel schedules. The proposed model is adaptable to other regions with similar travel surveys and emerging data sources, like call detail records, advancing the use of these data for activity-based models in a cost-effective, easily updated manner.

Sparse Autoencoders (SAEs) have emerged as a powerful unsupervised method for extracting sparse representations from language models, yet scalable training remains a significant challenge. We introduce a suite of 256 SAEs, trained on each layer and sublayer of the Llama-3.1-8B-Base model, with 32K and 128K features. Modifications to a state-of-the-art SAE variant, Top-K SAEs, are evaluated across multiple dimensions. In particular, we assess the generalizability of SAEs trained on base models to longer contexts and fine-tuned models. Additionally, we analyze the geometry of learned SAE latents, confirming that \emph{feature splitting} enables the discovery of new features. The Llama Scope SAE checkpoints are publicly available at~\url{https://huggingface.co/fnlp/Llama-Scope}, alongside our scalable training, interpretation, and visualization tools at \url{https://github.com/OpenMOSS/Language-Model-SAEs}. These contributions aim to advance the open-source Sparse Autoencoder ecosystem and support mechanistic interpretability research by reducing the need for redundant SAE training., Comment: 22pages, 12 figures

The Large Magellanic Cloud (LMC) is the nearest massive galaxy to the Milky Way. Its circumgalactic medium is complex and multi-phase, containing both stripped HI structures like the Magellanic Stream and Bridge, and a diffuse warm corona seen in high-ion absorption. We analyze 28 AGN sightlines passing within 35 kpc of the LMC with archival HST/COS spectra to characterize the cool (T\approx10^4$ K) gas in the LMC CGM, via new measurements of UV absorption in six low ions (OI, FeII, SiII, AlII, SII, and NiII) and one intermediate ion (SiIII). We show that a declining column-density profile is present in all seven ions, with the low-ion profiles having a steeper slope than the high-ion profiles in CIV and SiIV reported by Krishnarao et al. 2022. Crucially, absorption at the LMC systemic velocity is only detected (in all ions) out to 17 kpc. Beyond this distance, the gas has a lower velocity and is associated with the Magellanic Stream. These results demonstrate that the LMC's CGM is composed of two distinct components: a compact inner halo extending to 17 kpc, and a more extended stripped region associated with the Stream. The compactness and truncation of the LMC's inner CGM agree with recent simulations of ram-pressure stripping of the LMC by the Milky Way's extended corona., Comment: Accepted for publication in ApJL

The focused helium ion beam microscope is a versatile imaging and nanofabrication instrument enabling direct-write lithography with sub-10-nm resolution. Subsurface damage and swelling of substrates due to helium ion implantation is generally unwanted. However, these effects can also be leveraged for specific nanofabrication tasks. To explore this, we investigate focused helium ion beam induced swelling of bulk crystalline silicon and free-standing amorphous silicon nitride membranes using various irradiation strategies. We show that the creation of near-surface voids due to helium ion implantation can be used to induce surface nanostructure and create subsurface nanochannels. By tailoring the ion dose and beam energy, the size and depth of the swollen features can be controlled. Swelling heights of several hundred nanometers are demonstrated and for the embedded nanochannels, void internal diameters down to 30nm are shown. Potential applications include the engineering of texturized substrates and the prototyping of on-chip nanofluidic transport devices.

The resolution-line edge roughness-sensitivity tradeoff has motivated exploration of potential improvements using defined sequence polymers and polymer-bound photoacid generators and quenchers. In this study we characterize the internal structures of positive tone photoresist polymer films formed from defined sequence polymers and compare them to random copolymers of the same composition. We model their imaging to connect initial to developable film structures. We use a polymer packing algorithm to simulate films of diverse compositions and locations of photoacid generators and quenchers, using the composition of an ESCAP photoresist. We use a simple EUV exposure-deprotection algorithm to model developable image formation within them. In all cases, the spatial distribution of chemical moieties in the film for defined sequence polymers is nearly indistinguishable from random copolymers. We evaluate several exposure-deprotection scenarios, and find that a defined sequence copolymer has a distinctive developable image under certain circumstances. The use of defined sequence polymers within a photoresist layer does not automatically result in improved imaging, however they do have some characteristics different from random polymers of the same composition. Further study of these characteristics may provide a route to improved control over the nanoscale imaging process.

The recent JWST detections of carbon-bearing molecules in a habitable-zone sub-Neptune have opened a new era in the study of low-mass exoplanets. The sub-Neptune regime spans a wide diversity of planetary interiors and atmospheres not witnessed in the solar system, including mini-Neptunes, super-Earths, and water worlds. Recent works have investigated the possibility of gas dwarfs, with rocky interiors and thick H$_2$-rich atmospheres, to explain aspects of the sub-Neptune population, including the radius valley. Interactions between the H$_2$-rich envelope and a potential magma ocean may lead to observable atmospheric signatures. We report a coupled interior-atmosphere modelling framework for gas dwarfs to investigate the plausibility of magma oceans on such planets and their observable diagnostics. We find that the surface-atmosphere interactions and atmospheric composition are sensitive to a wide range of parameters, including the atmospheric and internal structure, mineral composition, volatile solubility and atmospheric chemistry. While magma oceans are typically associated with high-temperature rocky planets, we assess if such conditions may be admissible and observable for temperate sub-Neptunes. We find that a holistic modelling approach is required for this purpose and to avoid unphysical model solutions. We find using our model framework and considering the habitable-zone sub-Neptune K2-18 b as a case study that its observed atmospheric composition is incompatible with a magma ocean scenario. We identify key atmospheric molecular and elemental diagnostics, including the abundances of CO$_2$, CO, NH$_3$ and, potentially, S-bearing species. Our study also underscores the need for fundamental material properties for accurate modelling of such planets., Comment: Accepted for publication in ApJ, 30 pages, 14 figures

Although fundamental to the advancement of Machine Learning, the classic evaluation metrics extracted from the confusion matrix, such as precision and F1, are limited. Such metrics only offer a quantitative view of the models' performance, without considering the complexity of the data or the quality of the hit. To overcome these limitations, recent research has introduced the use of psychometric metrics such as Item Response Theory (IRT), which allows an assessment at the level of latent characteristics of instances. This work investigates how IRT concepts can enrich a confusion matrix in order to identify which model is the most appropriate among options with similar performance. In the study carried out, IRT does not replace, but complements classical metrics by offering a new layer of evaluation and observation of the fine behavior of models in specific instances. It was also observed that there is 97% confidence that the score from the IRT has different contributions from 66% of the classical metrics analyzed., Comment: 15 pages, 8 figures, 3 tables, submitted for the BRACIS'24 conference

Gate quantum computers generate significant interest due to their potential to solve certain difficult problems such as prime factorization in polynomial time. Computer vision researchers have long been attracted to the power of quantum computers. Robust fitting, which is fundamentally important to many computer vision pipelines, has recently been shown to be amenable to gate quantum computing. The previous proposed solution was to compute Boolean influence as a measure of outlyingness using the Bernstein-Vazirani quantum circuit. However, the method assumed a quantum implementation of an $\ell_\infty$ feasibility test, which has not been demonstrated. In this paper, we take a big stride towards quantum robust fitting: we propose a quantum circuit to solve the $\ell_\infty$ feasibility test in the 1D case, which allows to demonstrate for the first time quantum robust fitting on a real gate quantum computer, the IonQ Aria. We also show how 1D Boolean influences can be accumulated to compute Boolean influences for higher-dimensional non-linear models, which we experimentally validate on real benchmark datasets., Comment: Accepted by the European Conference on Computer Vision 2024 (ECCV2024) as Oral. The paper is written for a computer vision audience who generally has minimal quantum physics background

Nanopores in 2D materials are of significant interest in advanced membrane technologies aimed at the sensing and separation of ions and molecules. These applications necessitate 2D nanopores that are precise in size and shape, and abundant in number. However, conventional fabrication techniques often struggle to achieve both high precision and throughput. In this study, we introduce a decoupled seeding and growth approach designed to overcome this limitation. The method allows the controlled fabrication of ensembles of nanopores with narrow size distribution and is demonstrated for free-standing monolayer hexagonal boron nitride. Using light ion showering, we first create vacancy defect seeds. These seeds are then expanded into triangular nanopores through element-specific preferential atom removal under broad-beam electron irradiation in a transmission electron microscope. Nanopore density and size are controlled by the ion and electron irradiation doses, respectively. During the electron irradiation step, high-resolution imaging allows real-time tracking of the nanopore formation process, enabling the highest level of control. An additional workflow is introduced using thermal annealing in air for the nanopore growth step, delivering the most flexible platform for nanopore fabrication over larger areas with the significant benefit of concurrently removing surface hydrocarbon contamination to mitigate pore clogging and distortion. This study provides researchers with a novel approach to create ensembles of meticulously designed nanopores with well-controlled size and geometry, facilitating the development of next-generation membrane devices that will demand high precision and high throughput nanopore fabrication pipelines.

Surrogate models that combine dimensionality reduction and regression techniques are essential to reduce the need for costly high-fidelity computational fluid dynamics data. New approaches using $\beta$-Variational Autoencoder ($\beta$-VAE) architectures have shown promise in obtaining high-quality low-dimensional representations of high-dimensional flow data while enabling physical interpretation of their latent spaces. We propose a surrogate model based on latent space regression to predict pressure distributions on a transonic wing given the flight conditions: Mach number and angle of attack. The $\beta$-VAE model, enhanced with Principal Component Analysis (PCA), maps high-dimensional data to a low-dimensional latent space, showing a direct correlation with flight conditions. Regularization through $\beta$ requires careful tuning to improve overall performance, while PCA preprocessing helps to construct an effective latent space, improving autoencoder training and performance. Gaussian Process Regression is used to predict latent space variables from flight conditions, showing robust behavior independent of $\beta$, and the decoder reconstructs the high-dimensional pressure field data. This pipeline provides insight into unexplored flight conditions. Furthermore, a fine-tuning process of the decoder further refines the model, reducing the dependence on $\beta$ and enhancing accuracy. Structured latent space, robust regression performance, and significant improvements in fine-tuning collectively create a highly accurate and efficient surrogate model. Our methodology demonstrates the effectiveness of $\beta$-VAEs for aerodynamic surrogate modeling, offering a rapid, cost-effective, and reliable alternative for aerodynamic data prediction., Comment: 19 pages, 12 figures

Citizen science offers an opportunity for ordinary people, known as citizen scientists or citizen astronomers in the context of astronomy, to contribute to scientific research. The Pan-African Citizen Science e-Lab (PACS e-Lab) was founded to promote and engage the African public in citizen science and soft astronomy research to advance space research and exploration and enhance space education and outreach. PACS e-Lab, in collaboration with several international astronomy research, education, and outreach organizations, currently runs several projects including but not limited to asteroid search, exoplanet photometry, research writing for peer-reviewed publications, astrophoto visual development, and Amateur Radio contact with astronauts aboard the International Space Station (ARISS). Despite several challenges, the group has engaged over 600 Africans from more than 40 countries and is working towards covering the entire continent in the future. PACS e-Lab's development efforts resonate with seven United Nations Sustainable Development Goals (UN-SDGs)., Comment: 32 pages,17 figures, 2 Tables

Asteroid search is a global effort for planetary defense. The International Astronomical Search Collaboration (IASC) is the leading global educational outreach program that provides high-quality astronomical datasets to citizen scientists to discover asteroids. Since December 4, 2020, the Pan-African Citizen Science e-Lab (PACS e-Lab) has been IASC's biggest partner on the continent in recruiting and training citizen scientists in asteroid search endeavors. Over 30 asteroids have been discovered by 60 citizen scientists. About 595 citizen scientists from over 40 countries have been engaged in the project up to the time of composing this literature. The group is set to expand its endeavors to the rest of the continent and aims to engage thousands of citizen scientists., Comment: 16 pages, 10 figures, 5 Tables

Enzymes are important proteins that catalyze chemical reactions. In recent years, machine learning methods have emerged to predict enzyme function from sequence; however, there are no standardized benchmarks to evaluate these methods. We introduce CARE, a benchmark and dataset suite for the Classification And Retrieval of Enzymes (CARE). CARE centers on two tasks: (1) classification of a protein sequence by its enzyme commission (EC) number and (2) retrieval of an EC number given a chemical reaction. For each task, we design train-test splits to evaluate different kinds of out-of-distribution generalization that are relevant to real use cases. For the classification task, we provide baselines for state-of-the-art methods. Because the retrieval task has not been previously formalized, we propose a method called Contrastive Reaction-EnzymE Pretraining (CREEP) as one of the first baselines for this task and compare it to the recent method, CLIPZyme. CARE is available at https://github.com/jsunn-y/CARE/.

The successful use of molecular dyes for solar energy conversion requires efficient charge injection, which in turn requires the formation of states with sufficiently long lifetimes (e.g. triplets). The molecular structure elements that confer this property can be found empirically, however computational predictions using $\textit{ab initio}$ electronic structure methods are invaluable to identify structure-property relations for dye sensitizers. The primary challenge for simulations to elucidate the electronic and nuclear origins of these properties is a spin-orbit interaction which drives transitions between electronic states. In this work, we present a computational analysis of the spin-orbit corrected linear absorption cross sections and intersystem crossing rate coefficients for a derivative set of phosphonated tris(2,2'-bipyridine)ruthenium(2+) dye molecules. After sampling the ground state vibrational distributions, the predicted linear absorption cross sections indicate that the mixture between singlet and triplet states plays a crucial role in defining the line shape of the metal-to-ligand charge transfer bands in these derivatives. Additionally, an analysis of the intersystem crossing rate coefficients suggests that transitions from the singlet into the triplet manifolds are ultrafast with rate coefficients on the order of $10^{13}$ s$^{-1}$ for each dye molecule., Comment: 38 pages, 7 figures

Does the cross-product of the position and the electromagnetic momentum density, g, include the optical spin momentum? We answer this long-standing question in the affirmative by evaluating, explicitly, the torque exerted on a particle by a beam of light carrying angular momentum.

Vagus Nerve Stimulation (VNS) is an established palliative treatment for drug resistant epilepsy. While effective for many patients, its mechanism of action is incompletely understood. Predicting individuals' response, or optimum stimulation parameters, is challenging. Computational modelling has informed other problems in epilepsy but, to our knowledge, has not been applied to VNS. We started with an established, four-population neural mass model (NMM), capable of reproducing the seizure-like dynamics of a thalamocortical circuit. We extended this to include 18 further neural populations, representing nine other brain regions relevant to VNS, with connectivity based on existing literature. We modelled stimulated afferent vagal fibres as projecting to the nucleus tractus solitarius (NTS), which receives input from the vagus nerve in vivo. Bifurcation analysis of a deterministic version of the model showed higher background NTS input made the model monostable at a fixed point (FP), representing normal activity, while lower inputs produce bistability between the FP and a limit cycle (LC), representing the seizure state. Adding noise produced transitions between seizure and normal states. This stochastic model spent decreasing time in the seizure state with increasing background NTS input, until seizures were abolished, consistent with the deterministic model. Simulated VNS stimulation, modelled as a 30 Hz square wave, was summed with the background input to the NTS and was found to reduce total seizure duration in a dose-dependent manner, similar to expectations in vivo. We have successfully produced an in silico model of VNS in epilepsy, capturing behaviour seen in vivo. This may aid understanding therapeutic mechanisms of VNS in epilepsy and provides a starting point to (i) determine which patients might respond best to VNS, and (ii) optimise individuals' treatments.

Two-dimensional van der Waals magnets hosting topological magnetic textures, such as skyrmions, show promise for applications in spintronics and quantum computing. Electrical control of these topological spin textures would enable novel devices with enhanced performance and functionality. Here, using electron microscopy combined with in situ electric and magnetic biasing, we show that the skyrmion chirality, whether left-handed or right-handed, in insulating Cr2Ge2Te6, is controlled by external electric field direction applied during magnetic field cooling process. The electric-field-tuned chirality remains stable, even amid variations in magnetic and electric fields. Our theoretical investigation reveals that nonzero Dzyaloshinskii-Moriya interactions between the nearest neighbors, induced by the external electric field, change their sign upon reversing the electric field direction, thereby facilitating chirality selection. The electrical control of magnetic chirality demonstrated in this study can be extended to other non-metallic centrosymmetric skyrmion-hosting magnets, opening avenues for future device designs in topological spintronics and quantum computing.

This paper presents a novel simplification calculus for propositional logic derived from Peirce's existential graphs' rules of inference and implication graphs. Our rules can be applied to propositional logic formulae in nested form, are equivalence-preserving, guarantee a monotonically decreasing number of variables, clauses and literals, and maximise the preservation of structural problem information. Our techniques can also be seen as higher-level SAT preprocessing, and we show how one of our rules (TWSR) generalises and streamlines most of the known equivalence-preserving SAT preprocessing methods. In addition, we propose a simplification procedure based on the systematic application of two of our rules (EPR and TWSR) which is solver-agnostic and can be used to simplify large Boolean satisfiability problems and propositional formulae in arbitrary form, and we provide a formal analysis of its algorithmic complexity in terms of space and time. Finally, we show how our rules can be further extended with a novel n-ary implication graph to capture all known equivalence-preserving preprocessing procedures., Comment: 19 pages, 12 figures. Under consideration in Theory and Practice of Logic Programming (TPLP)

Providing external knowledge to Large Language Models (LLMs) is a key point for using these models in real-world applications for several reasons, such as incorporating up-to-date content in a real-time manner, providing access to domain-specific knowledge, and contributing to hallucination prevention. The vector database-based Retrieval Augmented Generation (RAG) approach has been widely adopted to this end. Thus, any part of external knowledge can be retrieved and provided to some LLM as the input context. Despite RAG approach's success, it still might be unfeasible for some applications, because the context retrieved can demand a longer context window than the size supported by LLM. Even when the context retrieved fits into the context window size, the number of tokens might be expressive and, consequently, impact costs and processing time, becoming impractical for most applications. To address these, we propose CRAG, a novel approach able to effectively reduce the number of prompting tokens without degrading the quality of the response generated compared to a solution using RAG. Through our experiments, we show that CRAG can reduce the number of tokens by at least 46\%, achieving more than 90\% in some cases, compared to RAG. Moreover, the number of tokens with CRAG does not increase considerably when the number of reviews analyzed is higher, unlike RAG, where the number of tokens is almost 9x higher when there are 75 reviews compared to 4 reviews.

In a context of a continuous digitalisation of processes, organisations must deal with the challenge of detecting anomalies that can reveal suspicious activities upon an increasing volume of data. To pursue this goal, audit engagements are carried out regularly, and internal auditors and purchase specialists are constantly looking for new methods to automate these processes. This work proposes a methodology to prioritise the investigation of the cases detected in two large purchase datasets from real data. The goal is to contribute to the effectiveness of the companies' control efforts and to increase the performance of carrying out such tasks. A comprehensive Exploratory Data Analysis is carried out before using unsupervised Machine Learning techniques addressed to detect anomalies. A univariate approach has been applied through the z-Score index and the DBSCAN algorithm, while a multivariate analysis is implemented with the k-Means and Isolation Forest algorithms, and the Silhouette index, resulting in each method having a transaction candidates' proposal to be reviewed. An ensemble prioritisation of the candidates is provided jointly with a proposal of explicability methods (LIME, Shapley, SHAP) to help the company specialists in their understanding.

In this article, we propose the detection of crowd anomalies through the extraction of information in the form of time series from video format using a multimodal approach. Through pattern recognition algorithms and segmentation, informative measures of the number of people and image occupancy are extracted at regular intervals, which are then analyzed to obtain trends and anomalous behaviors. Specifically, through temporal decomposition and residual analysis, intervals or specific situations of unusual behaviors are identified, which can be used in decision-making and improvement of actions in sectors related to human movement such as tourism or security. The application of this methodology on the webcam of Turisme Comunitat Valenciana in the town of Morella (Comunitat Valenciana, Spain) has provided excellent results. It is shown to correctly detect specific anomalous situations and unusual overall increases during the previous weekend and during the festivities in October 2023. These results have been obtained while preserving the confidentiality of individuals at all times by using measures that maximize anonymity, without trajectory recording or person recognition.

Collective intelligence of diverse groups is key for tackling many of today's grand challenges such as fostering resilience and climate adaptation. Information exchange across such diverse groups is crucial for collective intelligence, especially in volatile environments. To facilitate inter-group information exchange, Informational Boundary Spanners (IBSs) as pivotal information exchange 'hubs' are promising. However, the mechanisms that drive the emergence of IBSs remain poorly understood. To address this gap there is first a need for a method to identify and measure the emergence of IBSs. Second, an Agent-Based Modelling (ABM) framework is not available to systematically study mechanisms for the emergence of IBSs in volatile environments. Third, even though the ability to learn who provides high-quality information is thought to be essential to explain the emergence of IBSs, a rigorous test of this mechanism is missing. The learning mechanism is formalized using an ABM framework, with the model's outputs analyzed using the proposed IBS emergence measurement method. To illustrate both the method and the learning mechanism, we present a case study focused on information sharing in the volatile environment of a disaster. The study shows that learning constitutes a mechanism for the emergence of effective IBSs in (a) low-volatility environments characterised by low uncertainty and (b) in high-volatility environments characterised by rapid change if the number of inter-group connections is sufficient. With the method and model, this paper aims to lay the foundations for exploring mechanisms for the emergence of IBSs that facilitate inter-group information exchange. This article advances collective intelligence by providing the essential elements for measuring and understanding the emergence of IBSs and exploring the effect of learning on their emergence in volatile environments.

Predictive maintenance in manufacturing industry applications is a challenging research field. Packaging machines are widely used in a large number of logistic companies' warehouses and must be working uninterruptedly. Traditionally, preventive maintenance strategies have been carried out to improve the performance of these machines. However, this kind of policies does not take into account the information provided by the sensors implemented in the machines. This paper presents an expert system for the automatic estimation of work orders to implement predictive maintenance policies for packaging machines. The key idea is that, from a set of alarms related to sensors implemented in the machine, the expert system should take a maintenance action while optimizing the response time. The work order estimator will act as a classifier, yielding a binary decision of whether a machine must undergo a maintenance action by a technician or not, followed by an unsupervised anomaly detection-based filtering stage to audit the classifier's output. The methods used for anomaly detection were: One-Class Support Vector Machine (OCSVM), Minimum Covariance Determinant (MCD) and a majority (hard) voting ensemble of them. All anomaly detection methods improve the performance of the baseline classifer but the best performance in terms of F1 score was obtained by the majority voting ensemble.

Background: Understanding healthy human brain function is crucial to identify and map pathological tissue within it. Whilst previous studies have mapped intracranial EEG (icEEG) from non-epileptogenic brain regions, these maps do not consider the effects of age and sex. Further, most existing work on icEEG has often suffered from a small sample size due to the modality's invasive nature. Here, we substantially increase the subject sample size compared to existing literature, to create a multi-centre, normative map of brain activity which additionally considers the effects of age, sex and recording hospital. Methods: Using interictal icEEG recordings from n = 502 subjects originating from 15 centres, we constructed a normative map of non-pathological brain activity by regressing age and sex on relative band power in five frequency bands, whilst accounting for the hospital effect. Results: Recording hospital significantly impacted normative icEEG maps in all frequency bands, and age was a more influential predictor of band power than sex. The age effect varied by frequency band, but no spatial patterns were observed at the region-specific level. Certainty about regression coefficients was also frequency band specific and moderately impacted by sample size. Conclusion: The concept of a normative map is well-established in neuroscience research and particularly relevant to the icEEG modality, which does not allow healthy control baselines. Our key results regarding the hospital site and age effect guide future work utilising normative maps in icEEG.

The Hyperspectral Thermal Imager (HyTI) is a technology demonstration mission that will obtain high spatial, spectral, and temporal resolution long-wave infrared images of Earth's surface from a 6U cubesat. HyTI science requires that the pointing accuracy of the optical axis shall not exceed 2.89 arcsec over the 0.5 ms integration time due to microvibration effects (known as jitter). Two sources of vibration are a cryocooler that is added to maintain the detector at 68 K and three orthogonally placed reaction wheels that are a part of the attitude control system. Both of these parts will introduce vibrations that are propagated through to the satellite structure while imaging. Typical methods of characterizing and measuring jitter involve complex finite element methods and specialized equipment and setups. In this paper, we describe a novel method of characterizing jitter for small satellite systems that is low-cost and minimally modifies the subject's mass distribution. The metrology instrument is comprised of a laser source, a small mirror mounted via a 3D printed clamp to a jig, and a lateral effect position-sensing detector. The position-sensing detector samples 1000 Hz and can measure displacements as little as 0.15 arcsec at distances of one meter. This paper provides an experimental procedure that incrementally analyzes vibratory sources to establish causal relationships between sources and the vibratory modes they create. We demonstrate the capabilities of this metrology system and testing procedure on HyTI in the Hawaii Space Flight Lab's clean room. Results include power spectral density plots that show fundamental and higher-order vibratory modal frequencies. Results from metrology show that jitter from reaction wheels meets HyTI system requirements within 3$\sigma$., Comment: Accepted for the 2024 IEEE Aerospace Conference Proceedings

Mars, Venus and Earth are expected to have started in a hot molten state. Here, we discuss how these three terrestrial planets diverged in their evolution and what mechanisms could be the cause. We discuss that early-on after magma ocean crystallization the mantle/surface redox state and water inventory may already differ considerably, depending on planetary mass and orbital distance from the Sun. During the subsequent internal evolution, the three planets also diverged in terms of their tectonic regime, affecting the long-term planetary evolution via heat flux and outgassing rate, and possibly the physical state of their respective core and the onset and end of a planetary dynamo. We discuss how, throughout the evolution of these rocky planets, the dominant process for atmospheric loss would shift from hydrodynamic escape and impact erosion to non-thermal escape, where small terrestrial planets like Mars are here more vulnerable to volatile loss., Comment: Accepted for publication in Treatise on Geochemistry, 3rd

Code-switching is a prevalent linguistic phenomenon in which multilingual individuals seamlessly alternate between languages. Despite its widespread use online and recent research trends in this area, research in code-switching presents unique challenges, primarily stemming from the scarcity of labelled data and available resources. In this study we investigate how pre-trained Language Models handle code-switched text in three dimensions: a) the ability of PLMs to detect code-switched text, b) variations in the structural information that PLMs utilise to capture code-switched text, and c) the consistency of semantic information representation in code-switched text. To conduct a systematic and controlled evaluation of the language models in question, we create a novel dataset of well-formed naturalistic code-switched text along with parallel translations into the source languages. Our findings reveal that pre-trained language models are effective in generalising to code-switched text, shedding light on the abilities of these models to generalise representations to CS corpora. We release all our code and data including the novel corpus at https://github.com/francesita/code-mixed-probes., Comment: Accepted for publication at Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). Data and code available at https://github.com/francesita/code-mixed-probes

Diffusion processes govern fundamental phenomena such as phase transformations, doping, and intercalation in van der Waals (vdW) bonded materials. Here, we quantify the diffusion dynamics of W atoms by visualizing the motion of individual atoms at three different vdW interfaces: BN/vacuum, BN/BN, and BN/WSe2, by recording scanning transmission electron microscopy movies. Supported by density functional theory calculations, we infer that in all cases diffusion is governed by intermittent trapping at electron beam-generated defect sites. This leads to diffusion properties that depend strongly on the number of defects. These results suggest that diffusion and intercalation processes in vdW materials are highly tunable and sensitive to crystal quality. The demonstration of imaging, with high spatial and temporal resolution, of layers and individual atoms inside vdW heterostructures offers possibilities for direct visualization of diffusion and atomic interactions, as well as for experiments exploring atomic structures, their in-situ modification, and electrical property measurements of active devices combined with atomic resolution imaging., Comment: 12 pages, 6 figures

Automated phenotyping of plants for breeding and plant studies promises to provide quantitative metrics on plant traits at a previously unattainable observation frequency. Developers of tools for performing high-throughput phenotyping are, however, constrained by the availability of relevant datasets on which to perform validation. To this end, we present a spatio-temporal dataset of 3D point clouds of strawberry plants for two varieties, totalling 84 individual point clouds. We focus on the end use of such tools - the extraction of biologically relevant phenotypes - and demonstrate a phenotyping pipeline on the dataset. This comprises of the steps, including; segmentation, skeletonisation and tracking, and we detail how each stage facilitates the extraction of different phenotypes or provision of data insights. We particularly note that assessment is focused on the validation of phenotypes, extracted from the representations acquired at each step of the pipeline, rather than singularly focusing on assessing the representation itself. Therefore, where possible, we provide \textit{in silico} ground truth baselines for the phenotypes extracted at each step and introduce methodology for the quantitative assessment of skeletonisation and the length trait extracted thereof. This dataset contributes to the corpus of freely available agricultural/horticultural spatio-temporal data for the development of next-generation phenotyping tools, increasing the number of plant varieties available for research in this field and providing a basis for genuine comparison of new phenotyping methodology.

A digital twin is a virtual replica of a real-world physical phenomena that uses mathematical modeling to characterize and simulate its defining features. By constructing digital twins for disease processes, we can perform in-silico simulations that mimic patients' health conditions and counterfactual outcomes under hypothetical interventions in a virtual setting. This eliminates the need for invasive procedures or uncertain treatment decisions. In this paper, we propose a method to identify digital twin model parameters using only noninvasive patient health data. We approach the digital twin modeling as a composite inverse problem, and observe that its structure resembles pretraining and finetuning in self-supervised learning (SSL). Leveraging this, we introduce a physics-informed SSL algorithm that initially pretrains a neural network on the pretext task of learning a differentiable simulator of a physiological process. Subsequently, the model is trained to reconstruct physiological measurements from noninvasive modalities while being constrained by the physical equations learned in pretraining. We apply our method to identify digital twins of cardiac hemodynamics using noninvasive echocardiogram videos, and demonstrate its utility in unsupervised disease detection and in-silico clinical trials.

We consider the problem of graph searching with prediction recently introduced by Banerjee et al. (2022). In this problem, an agent, starting at some vertex $r$ has to traverse a (potentially unknown) graph $G$ to find a hidden goal node $g$ while minimizing the total distance travelled. We study a setting in which at any node $v$, the agent receives a noisy estimate of the distance from $v$ to $g$. We design algorithms for this search task on unknown graphs. We establish the first formal guarantees on unknown weighted graphs and provide lower bounds showing that the algorithms we propose have optimal or nearly-optimal dependence on the prediction error. Further, we perform numerical experiments demonstrating that in addition to being robust to adversarial error, our algorithms perform well in typical instances in which the error is stochastic. Finally, we provide alternative simpler performance bounds on the algorithms of Banerjee et al. (2022) for the case of searching on a known graph, and establish new lower bounds for this setting., Comment: AISTATS 2024

Recent studies have suggested the possibility of Hycean worlds, characterised by deep liquid water oceans beneath H$_2$-rich atmospheres. These planets significantly widen the range of planetary properties over which habitable conditions could exist. We conduct internal structure modelling of Hycean worlds to investigate the range of interior compositions, ocean depths and atmospheric mass fractions possible. Our investigation explicitly considers habitable oceans, where the surface conditions are limited to those that can support potential life. The ocean depths depend on the surface gravity and temperature, confirming previous studies, and span 10s to $\sim$1000 km for Hycean conditions, reaching ocean base pressures up to $\sim$6$\times$10$^4$ bar before transitioning to high-pressure ice. We explore in detail test cases of five Hycean candidates, placing constraints on their possible ocean depths and interior compositions based on their bulk properties. We report limits on their atmospheric mass fractions admissible for Hycean conditions, as well as those allowed for other possible interior compositions. For the Hycean conditions considered, across these candidates we find the admissible mass fractions of the H/He envelopes to be $\lesssim$10$^{-3}$. At the other extreme, the maximum H/He mass fractions allowed for these planets can be up to $\sim$4-8$\%$, representing purely rocky interiors with no H$_2$O layer. These results highlight the diverse conditions possible among these planets and demonstrate their potential to host habitable conditions under vastly different circumstances to the Earth. Upcoming JWST observations of candidate Hycean worlds will allow for improved constraints on the nature of their atmospheres and interiors., Comment: Accepted for publication in MNRAS

In this paper, we apply quasi-Monte Carlo (QMC) methods with an initial preintegration step to estimate cumulative distribution functions and probability density functions in uncertainty quantification (UQ). The distribution and density functions correspond to a quantity of interest involving the solution to an elliptic partial differential equation (PDE) with a lognormally distributed coefficient and a normally distributed source term. There is extensive previous work on using QMC to compute expected values in UQ, which have proven very successful in tackling a range of different PDE problems. However, the use of QMC for density estimation applied to UQ problems will be explored here for the first time. Density estimation presents a more difficult challenge compared to computing the expected value due to discontinuities present in the integral formulations of both the distribution and density. Our strategy is to use preintegration to eliminate the discontinuity by integrating out a carefully selected random parameter, so that QMC can be used to approximate the remaining integral. First, we establish regularity results for the PDE quantity of interest that are required for smoothing by preintegration to be effective. We then show that an $N$-point lattice rule can be constructed for the integrands corresponding to the distribution and density, such that after preintegration the QMC error is of order $\mathcal{O}(N^{-1+\epsilon})$ for arbitrarily small $\epsilon>0$. This is the same rate achieved for computing the expected value of the quantity of interest. Numerical results are presented to reaffirm our theory.

Pre-trained large language models, such as ChatGPT, archive outstanding performance in various reasoning tasks without supervised training and were found to have outperformed crowdsourcing workers. Nonetheless, ChatGPT's performance in the task of implicit discourse relation classification, prompted by a standard multiple-choice question, is still far from satisfactory and considerably inferior to state-of-the-art supervised approaches. This work investigates several proven prompting techniques to improve ChatGPT's recognition of discourse relations. In particular, we experimented with breaking down the classification task that involves numerous abstract labels into smaller subtasks. Nonetheless, experiment results show that the inference accuracy hardly changes even with sophisticated prompt engineering, suggesting that implicit discourse relation classification is not yet resolvable under zero-shot or few-shot settings., Comment: To appear at the Linguistic Annotation Workshop 2024

The sulfur dimer (S$_2$) is an important molecular constituent in cometary atmospheres and volcanic plumes on Jupiter's moon Io. It is also expected to play an important role in the photochemistry of exoplanets. The UV spectrum of S$_2$ contains transitions between vibronic levels above and below the dissociation limit, giving rise to a distinctive spectral signature. By using spectroscopic information from the literature, and the spectral simulation program PGOPHER, a UV line list of S$_2$ is provided. This line list includes the primary $B\,^{3}\Sigma^{-}_{u}-X\,^{3}\Sigma^{-}_{g}$ ($v'$=0-27, $v''$=0-10) electronic transition, where vibrational bands with $v'$$\geq$10 are predissociated. Intensities have been calculated from existing experimental and theoretical oscillator strengths, and semi-empirical strengths for the predissociated bands of S$_2$ have been derived from comparisons with experimental cross-sections. The S$_2$ line list also includes the $B''\,^{3}\Pi_{u}-X\,^{3}\Sigma^{-}_{g}$ ($v'$=0-19, $v''$=0-10) vibronic bands due to the strong interaction with the $B$ state. In summary, we present the new HITRAN-formatted S$_2$ line list and its validation against existing laboratory spectra. The extensive line list covers the spectral range 21700$-$41300~cm$^{-1}$ ($\sim$242$-$461~nm) and can be used for modeling both absorption and emission., Comment: Accepted to the Monthly Notices of the Royal Astronomical Society

We give a survey of the implosion construction, extending some of its aspects relating to hypertoric geometry from type $A$ to a general reductive group, and interpret it in the context of the Moore-Tachikawa category. We use these ideas to discuss how the contraction construction in symplectic geometry can be generalised to the hyperk\"ahler or complex symplectic situation., Comment: 23 Pages. To appear in Int. J. Math. (special issue in honour of Oscar Garcia Prada)