Search

The traditional and most common view of economists on the issue of (bad) uncertainty and its effects has been one of partial equilibrium. When the topic is approached from a macroeconomic perspective, the most frequent has been the examination of the effects of uncertainty shocks on short-term dynamics with various methods, but mainly with neo-Keynesian and statistical (vector autoregressive) models. This document responds to another concern and has two objectives: 1. to reflect on this issue with some instruments of the macroeconomist's toolbox related to a medium or long-term horizon, and 2. report a ciphering of the social cost of uncertainty in the Colombian case., Comment: 13 pages, 4 figures, 2 tables

The emission rate of atom-like photon sources can be significantly improved by coupling them to plasmonic resonant nanostructures. These arrangements function as nanoantennas, serving the dual purpose of enhancing light-matter interactions and decoupling the emitted photons. However, there is a contradiction between the requirements for these two tasks. A small resonator volume is necessary for maximizing interaction efficiency, while a large antenna mode volume is essential to achieve high emission directivity. In this work, we analyze a hybrid structure composed of a noble metal plasmonic resonant nanoparticle coupled to the atom-like emitter, designed to enhance the emission rate, alongside a corner reflector aimed at optimizing the angular distribution of the emitted photons. A comprehensive numerical study of silver and gold corner reflector nanoantennas, employing the finite-difference time-domain method, is presented. The results demonstrate that a well-designed corner reflector can significantly enhance photon emission directivity while also substantially boosting the emission rate.

The international collaboration designing and constructing the Deep Underground Neutrino Experiment (DUNE) at the Long-Baseline Neutrino Facility (LBNF) has developed a two-phase strategy toward the implementation of this leading-edge, large-scale science project. The 2023 report of the US Particle Physics Project Prioritization Panel (P5) reaffirmed this vision and strongly endorsed DUNE Phase I and Phase II, as did the European Strategy for Particle Physics. While the construction of the DUNE Phase I is well underway, this White Paper focuses on DUNE Phase II planning. DUNE Phase-II consists of a third and fourth far detector (FD) module, an upgraded near detector complex, and an enhanced 2.1 MW beam. The fourth FD module is conceived as a "Module of Opportunity", aimed at expanding the physics opportunities, in addition to supporting the core DUNE science program, with more advanced technologies. This document highlights the increased science opportunities offered by the DUNE Phase II near and far detectors, including long-baseline neutrino oscillation physics, neutrino astrophysics, and physics beyond the standard model. It describes the DUNE Phase II near and far detector technologies and detector design concepts that are currently under consideration. A summary of key R&D goals and prototyping phases needed to realize the Phase II detector technical designs is also provided. DUNE's Phase II detectors, along with the increased beam power, will complete the full scope of DUNE, enabling a multi-decadal program of groundbreaking science with neutrinos.

Rapid progress in general-purpose AI has sparked significant interest in "red teaming," a practice of adversarial testing originating in military and cybersecurity applications. AI red teaming raises many questions about the human factor, such as how red teamers are selected, biases and blindspots in how tests are conducted, and harmful content's psychological effects on red teamers. A growing body of HCI and CSCW literature examines related practices-including data labeling, content moderation, and algorithmic auditing. However, few, if any have investigated red teaming itself. Future studies may explore topics ranging from fairness to mental health and other areas of potential harm. We aim to facilitate a community of researchers and practitioners who can begin to meet these challenges with creativity, innovation, and thoughtful reflection., Comment: Updated with camera-ready version

Since the Transformer architecture emerged, language model development has grown, driven by their promising potential. Releasing these models into production requires properly understanding their behavior, particularly in sensitive domains like medicine. Despite this need, the medical literature still lacks practical assessment of pre-trained language models, which are especially valuable in settings where only consumer-grade computational resources are available. To address this gap, we have conducted a comprehensive survey of language models in the medical field and evaluated a subset of these for medical text classification and conditional text generation. The subset includes 53 models with 110 million to 13 billion parameters, spanning the Transformer-based model families and knowledge domains. Different approaches are employed for text classification, including zero-shot learning, enabling tuning without the need to train the model. These approaches are helpful in our target settings, where many users of language models find themselves. The results reveal remarkable performance across the tasks and datasets evaluated, underscoring the potential of certain models to contain medical knowledge, even without domain specialization. This study thus advocates for further exploration of model applications in medical contexts, particularly in computational resource-constrained settings, to benefit a wide range of users. The code is available on https://github.com/anpoc/Language-models-in-medicine.

This paper presents a strategy for efficient quantum circuit design for density estimation. The strategy is based on a quantum-inspired algorithm for density estimation and a circuit optimisation routine based on memetic algorithms. The model maps a training dataset to a quantum state represented by a density matrix through a quantum feature map. This training state encodes the probability distribution of the dataset in a quantum state, such that the density of a new sample can be estimated by projecting its corresponding quantum state onto the training state. We propose the application of a memetic algorithm to find the architecture and parameters of a variational quantum circuit that implements the quantum feature map, along with a variational learning strategy to prepare the training state. Demonstrations of the proposed strategy show an accurate approximation of the Gaussian kernel density estimation method through shallow quantum circuits illustrating the feasibility of the algorithm for near-term quantum hardware., Comment: 15 pages, 8 figures, presented at QTML 2023

Constrained optimization offers a powerful framework to prescribe desired behaviors in neural network models. Typically, constrained problems are solved via their min-max Lagrangian formulations, which exhibit unstable oscillatory dynamics when optimized using gradient descent-ascent. The adoption of constrained optimization techniques in the machine learning community is currently limited by the lack of reliable, general-purpose update schemes for the Lagrange multipliers. This paper proposes the $\nu$PI algorithm and contributes an optimization perspective on Lagrange multiplier updates based on PI controllers, extending the work of Stooke, Achiam and Abbeel (2020). We provide theoretical and empirical insights explaining the inability of momentum methods to address the shortcomings of gradient descent-ascent, and contrast this with the empirical success of our proposed $\nu$PI controller. Moreover, we prove that $\nu$PI generalizes popular momentum methods for single-objective minimization. Our experiments demonstrate that $\nu$PI reliably stabilizes the multiplier dynamics and its hyperparameters enjoy robust and predictable behavior., Comment: Published at ICML 2024. Code available at https://github.com/motahareh-sohrabi/nuPI

The present paper is devoted to a study of the equilibrium configurations of slowly rotating anisotropic stars in the framework of general relativity. For that purpose, we provide the equations of structure where the rotation is treated to second order in the angular velocity. These equations extend those first derived by Hartle for slowly rotating isotropic stars. As an application of the new formalism, we study the rotational properties of Bowers-Liang fluid spheres. A result of particular interest is that the ellipticity and mass quadrupole moment are negative for certain highly anisotropic configurations; thus, such systems are prolate rather than oblate. Furthermore, for configurations with high anisotropy and compactness close to their critical value, quantities like the moment of inertia, change of mass, and mass quadrupole moment approach to the corresponding Kerr black hole values, similar to other ultracompact systems like sub-Buchdahl Schwarzschild stars and analytic rotating gravastars., Comment: 27 pages, 30 figures. Discussion in Appendix B extended. Matches the published version in PRD

Many of the world's workers rely on digital platforms for their income. In Venezuela, a nation grappling with extreme inflation and where most of the workforce is self-employed, data production platforms for machine learning have emerged as a viable opportunity for many to earn a flexible income in US dollars. Platform workers are deeply interconnected within a vast network of firms and entities that act as intermediaries for wage payments in digital currencies and its subsequent conversion to the national currency, the bolivar. Past research on embeddedness has noted that being intertwined in multi-tiered socioeconomic networks of companies and individuals can offer significant rewards to social participants, while also connoting a particular set of limitations. This paper furnishes qualitative evidence regarding how this deep embeddedness impacts platform workers in Venezuela. Given the backdrop of a national crisis and rampant hyperinflation, the perks of receiving wages through various financial platforms include access to a more stable currency and the ability to save and invest outside the national financial system. However, relying on numerous digital and local intermediaries often diminishes income due to transaction fees. Moreover, this introduces heightened financial risks, particularly due to the unpredictable nature of cryptocurrencies as an investment. The over-reliance on external financial platforms erodes worker autonomy through power dynamics that lean in favor of the platforms that set the transaction rules and prices. These findings present a multifaceted perspective on deep embeddedness in platform labor, highlighting how the rewards of financial intermediation often come at a substantial cost for the workers in unstable situations, who are saddled with escalating financial risks.

Waveguide geometries combining bent asymmetric coupled structures and adiabatic transitions shaped as partial Euler bends are put forward and theoretically analyzed in this work. Designs aiming to reduce the radiation loss, with application in curved waveguide sections and Q-enhanced microresonators, and polarization selection geometries, both for the silicon nitride platform, are studied using highly accurate numerical techniques., Comment: Final published version

Room temperature photoluminescence and optical absorption spectra have been measured in wurtzite In$_{x}$Al$_{y}$Ga$_{1-x-y}$N (x $\approx$ 0.06, 0.02 < y < 0.27) layers grown by molecular beam epitaxy. Photoluminescence spectra show both an enhancement of the integrated intensity and an increasing Stokes shift with the Al content. Both effects arise from an Al-enhanced exciton localization revealed by the S- and W-shaped temperature dependences of the photoluminescence emission energy and bandwidth respectively. Present results point to these materials as a promising choice for the active region in efficient light emitters. An In-related bowing parameter of 1.6 eV was derived from optical absorption data.

The idea of implementing electroluminescence-based amplification through transparent multi-hole structures (FAT-GEMs) has been entertained for some time. Arguably, for such a technology to be attractive it should perform at least at a level comparable to conventional alternatives based on wires or meshes. We present now a detailed calorimetric study carried out for 5.9~keV X-rays in xenon, for pressures ranging from 2 to 10~bar, resorting to different geometries, production and post-processing techniques. At a reference voltage 5~times above the electroluminescence threshold ($E_{EL,th}\sim0.7$~kV/cm/bar), the number of photoelectrons measured for the best structure was found to be just 18\%~below that obtained for a double-mesh with the same thickness and at the same distance. The energy resolution stayed within 10\% (relative) of the double-mesh value. An innovative characteristic of the structure is that vacuum ultraviolet (VUV) transparency of the polymethyl methacrylate (PMMA) substrate was achieved, effectively, through tetraphenylbutadiene (TPB) coating of the electroluminescence channels combined with indium tin oxide (ITO) coating of the electrodes. This resulted in a $\times 2.25$-increased optical yield (compared to the bare structure), that was found to be in good agreement with simulations if assuming a TPB wavelength-shifting-efficiency at the level of WLSE=0.74-1.28, compatible with expected values. This result, combined with the stability demonstrated for the TPB coating under electric field (over 20~h of continuous operation), shows great potential to revolutionize electroluminescence-based instrumentation.

Spontaneous emission is one of the most fundamental out-of-equilibrium processes in which an excited quantum emitter relaxes to the ground state due to quantum fluctuations. In this process, a photon is emitted that can interact with other nearby emitters and establish quantum correlations between them, e.g., via super and subradiance effects. One way to modify these photon-mediated interactions is to alter the dipole radiation patterns of the emitter, e.g., by placing photonic crystals near them. One recent example is the generation of strong directional emission patterns-key to enhancing super and subradiance effects-in two dimensions by employing photonic crystals with band structures characterized by linear isofrequency contours and saddle-points. However, these studies have predominantly used oversimplified toy models, overlooking the electromagnetic field's intricacies in actual materials, including aspects like geometrical dependencies, emitter positions, and polarization. Our study delves into the interaction between these directional emission patterns and the aforementioned variables, revealing the untapped potential to fine-tune collective quantum optical phenomena., Comment: 10 pages and 6 Figures. Submmitted to Nanophotonics journal

Foundation models hold promise for transforming AI in healthcare by providing modular components that are easily adaptable to downstream healthcare tasks, making AI development more scalable and cost-effective. Structured EHR foundation models, trained on coded medical records from millions of patients, demonstrated benefits including increased performance with fewer training labels, and improved robustness to distribution shifts. However, questions remain on the feasibility of sharing these models across different hospitals and their performance for local task adaptation. This multi-center study examined the adaptability of a recently released structured EHR foundation model ($FM_{SM}$), trained on longitudinal medical record data from 2.57M Stanford Medicine patients. Experiments were conducted using EHR data at The Hospital for Sick Children and MIMIC-IV. We assessed both adaptability via continued pretraining on local data, and task adaptability compared to baselines of training models from scratch at each site, including a local foundation model. We evaluated the performance of these models on 8 clinical prediction tasks. In both datasets, adapting the off-the-shelf $FM_{SM}$ matched the performance of GBM models locally trained on all data while providing a 13% improvement in settings with few task-specific training labels. With continued pretraining on local data, label efficiency substantially improved, such that $FM_{SM}$ required fewer than 1% of training examples to match the fully trained GBM's performance. Continued pretraining was also 60 to 90% more sample-efficient than training local foundation models from scratch. Our findings show that adapting shared EHR foundation models across hospitals provides improved prediction performance at less cost, underscoring the utility of base foundation models as modular components to streamline the development of healthcare AI., Comment: 46 pages, 5 figures, 3 tables, 14 appendices

Model pruning is a popular approach to enable the deployment of large deep learning models on edge devices with restricted computational or storage capacities. Although sparse models achieve performance comparable to that of their dense counterparts at the level of the entire dataset, they exhibit high accuracy drops for some data sub-groups. Existing methods to mitigate this disparate impact induced by pruning (i) rely on surrogate metrics that address the problem indirectly and have limited interpretability; or (ii) scale poorly with the number of protected sub-groups in terms of computational cost. We propose a constrained optimization approach that directly addresses the disparate impact of pruning: our formulation bounds the accuracy change between the dense and sparse models, for each sub-group. This choice of constraints provides an interpretable success criterion to determine if a pruned model achieves acceptable disparity levels. Experimental results demonstrate that our technique scales reliably to problems involving large models and hundreds of protected sub-groups., Comment: Published at ICLR 2024. Code available at https://github.com/merajhashemi/balancing-act

We investigated the stability of the bottom boundary layer (BBL) beneath periodic internal solitary waves (ISWs) of depression over a flat bottom through two-dimensional direct numerical simulations. We explored the effects of variation in wave Reynolds number $Re_{ISW}$ and wave period on the nature of the instability, and energy production in the separated BBL. The instability characteristics and rate of vortex shedding of the BBL were strongly dependent on $Re_{ISW}$. The BBL was laminar and convectively unstable at $Re_{ISW}$ 90 and 300, respectively. At $Re_{ISW}=300$, the convective wave packet was periodically amplified by each successive ISW, until vortex-shedding occurred. This implies noise-amplification behavior and suggests that the discrepancies in the critical $Re_{ISW}$, for vortex shedding between lab and different numerical simulations, are due to differences in background seed noise. Instability energy decreased under the front shoulder of the ISW, analogous to flow relaminarization under a favourable pressure gradient. At larger $Re_{ISW}=900$, the BBL was initially convectively unstable, and then the instability tracked with the ISW, characteristic of global instability, regardless of the ISW periodicity. The simulated initial convective instability at both $Re_{ISW}$ 300 and 900 is in agreement with local linear stability analysis which predicts that the instability group speed is always lower than the ISW celerity. Increased free-stream perturbations and larger $Re_{ISW}$ shift the location of vortex shedding (and enhanced bed shear stress) closer to the ISW trough, thereby potentially changing the location of maximum sediment resuspension from the ISW, in agreement with field observations at higher $Re_{ISW}$.

Shampoo is an online and stochastic optimization algorithm belonging to the AdaGrad family of methods for training neural networks. It constructs a block-diagonal preconditioner where each block consists of a coarse Kronecker product approximation to full-matrix AdaGrad for each parameter of the neural network. In this work, we provide a complete description of the algorithm as well as the performance optimizations that our implementation leverages to train deep networks at-scale in PyTorch. Our implementation enables fast multi-GPU distributed data-parallel training by distributing the memory and computation associated with blocks of each parameter via PyTorch's DTensor data structure and performing an AllGather primitive on the computed search directions at each iteration. This major performance enhancement enables us to achieve at most a 10% performance reduction in per-step wall-clock time compared against standard diagonal-scaling-based adaptive gradient methods. We validate our implementation by performing an ablation study on training ImageNet ResNet50, demonstrating Shampoo's superiority over standard training recipes with minimal hyperparameter tuning., Comment: 38 pages, 8 figures, 5 tables

In$_2$O$_3$-based catalysts have shown high activity and selectivity for CO$_2$ hydrogenation to methanol, however the origin of the high performance of In$_2$O$_3$ is still unclear. To elucidate the initial steps of CO$_2$ hydrogenation over In$_2$O$_3$, we have combined X-ray Photoelectron Spectroscopy (XPS) and Density Functional Theory (DFT) calculations to study the adsorption of CO$_2$ on the In$_2$O$_3$(111) crystalline surface with different terminations, namely the stoichiometric, the reduced, and the hydroxylated surface, respectively. The combined approach confirms that the reduction of the surface results in the formation of In ad-atoms and that water dissociates on the surface at room temperature. A comparison of the experimental spectra and the computed core-level-shifts (using methanol and formic acid as benchmark molecules) suggests that CO$_2$ adsorbs as a carbonate on all surface terminations. We find that CO$_2$ adsorption is hindered by hydroxyl groups on the hydroxylated surface., Comment: 49 pages, 18 figures

We present a measurement of gravitational lensing over 1500 deg$^2$ of the Southern sky using SPT-3G temperature data at 95 and 150 GHz taken in 2018. The lensing amplitude relative to a fiducial Planck 2018 $\Lambda$CDM cosmology is found to be $1.020\pm0.060$, excluding instrumental and astrophysical systematic uncertainties. We conduct extensive systematic and null tests to check the robustness of the lensing measurements, and report a minimum-variance combined lensing power spectrum over angular multipoles of $50

Convolutional Neural Networks (CNN) have become a common choice for industrial quality control, as well as other critical applications in the Industry 4.0. When these CNNs behave in ways unexpected to human users or developers, severe consequences can arise, such as economic losses or an increased risk to human life. Concept extraction techniques can be applied to increase the reliability and transparency of CNNs through generating global explanations for trained neural network models. The decisive features of image datasets in quality control often depend on the feature's scale; for example, the size of a hole or an edge. However, existing concept extraction methods do not correctly represent scale, which leads to problems interpreting these models as we show herein. To address this issue, we introduce the Scale-Preserving Automatic Concept Extraction (SPACE) algorithm, as a state-of-the-art alternative concept extraction technique for CNNs, focused on industrial applications. SPACE is specifically designed to overcome the aforementioned problems by avoiding scale changes throughout the concept extraction process. SPACE proposes an approach based on square slices of input images, which are selected and then tiled before being clustered into concepts. Our method provides explanations of the models' decision-making process in the form of human-understandable concepts. We evaluate SPACE on three image classification datasets in the context of industrial quality control. Through experimental results, we illustrate how SPACE outperforms other methods and provides actionable insights on the decision mechanisms of CNNs. Finally, code for the implementation of SPACE is provided., Comment: 22 pages, 7 figures

In this work, the properties of the radiation emitted by a corner reflector with an electric dipole feeder are analyzed in the optical domain where the distance between the dipole and the corner apex can be large in terms of the optical wavelength. A comprehensive study of the fractal properties of the radiated intensity patterns is presented. The use of this setup for the realization of single-photon sources in photonic integrated circuits is also put forward and a detailed study of the emission properties of the device and its optimal configurations is presented., Comment: Final published version

We report the scintillation spectra of Ar-CF$_4$ mixtures in the range 210-800~nm, obtained under X-ray irradiation for various pressures (1-5~bar) and concentrations (0-100%). Special care was taken to eliminate effects related to space charge and recombination, so that results can be extrapolated following conventional wisdom to those expected for minimum ionizing particles under the typical electric fields employed in gaseous instrumentation. Our study sheds light into the microscopic pathways leading to scintillation in this family of mixtures., Comment: Updated to match current journal submission

The industry 4.0 is leveraging digital technologies and machine learning techniques to connect and optimize manufacturing processes. Central to this idea is the ability to transform raw data into human understandable knowledge for reliable data-driven decision-making. Convolutional Neural Networks (CNNs) have been instrumental in processing image data, yet, their ``black box'' nature complicates the understanding of their prediction process. In this context, recent advances in the field of eXplainable Artificial Intelligence (XAI) have proposed the extraction and localization of concepts, or which visual cues intervene on the prediction process of CNNs. This paper tackles the application of concept extraction (CE) methods to industry 4.0 scenarios. To this end, we modify a recently developed technique, ``Extracting Concepts with Local Aggregated Descriptors'' (ECLAD), improving its scalability. Specifically, we propose a novel procedure for calculating concept importance, utilizing a wrapper function designed for CNNs. This process is aimed at decreasing the number of times each image needs to be evaluated. Subsequently, we demonstrate the potential of CE methods, by applying them in three industrial use cases. We selected three representative use cases in the context of quality control for material design (tailored textiles), manufacturing (carbon fiber reinforcement), and maintenance (photovoltaic module inspection). In these examples, CE was able to successfully extract and locate concepts directly related to each task. This is, the visual cues related to each concept, coincided with what human experts would use to perform the task themselves, even when the visual cues were entangled between multiple classes. Through empirical results, we show that CE can be applied for understanding CNNs in an industrial context, giving useful insights that can relate to domain knowledge.