Search

Reinforcement Learning from Human Feedback (RLHF) is popular in large language models (LLMs), whereas traditional Reinforcement Learning (RL) often falls short. Current autonomous driving methods typically utilize either human feedback in machine learning, including RL, or LLMs. Most feedback guides the car agent's learning process (e.g., controlling the car). RLHF is usually applied in the fine-tuning step, requiring direct human "preferences," which are not commonly used in optimizing autonomous driving models. In this research, we innovatively combine RLHF and LLMs to enhance autonomous driving safety. Training a model with human guidance from scratch is inefficient. Our framework starts with a pre-trained autonomous car agent model and implements multiple human-controlled agents, such as cars and pedestrians, to simulate real-life road environments. The autonomous car model is not directly controlled by humans. We integrate both physical and physiological feedback to fine-tune the model, optimizing this process using LLMs. This multi-agent interactive environment ensures safe, realistic interactions before real-world application. Finally, we will validate our model using data gathered from real-life testbeds located in New Jersey and New York City.

A method for solving three dimensional discrete dislocation plasticity boundary-value problems using a monopole representation of the dislocations is presented. At each time step, the displacement, strain and stress fields in a finite body are obtained by superposition of infinite body dislocation fields and an image field that enforces the boundary conditions. The three dimensional infinite body fields are obtained by representing dislocations as being comprised of points, termed monopoles, that carry dislocation line and Burgers vector information. The image fields are obtained from a three dimensional linear elastic finite element calculation. The implementation of the coupling of the monopole representation with the finite element method, including the interaction of curved dislocations with free surfaces, is presented in some detail because it differs significantly from an implementation with a line based dislocation representation. Numerical convergence and the modeling of dislocation loop nucleation for large scale computations are investigated. The monopole discrete dislocation plasticity framework is used to investigate the effect of size and initial dislocation density on the torsion of wires with diameters varying over three orders of magnitude. Depending on the initial dislocation source density and the wire diameter, three regimes of torsion-twist response are obtained: (i) for wires with a sufficiently small diameter, plastic deformation is nucleation controlled and is strongly size dependent; (ii) for wires with larger diameters dislocation plasticity is dislocation interaction controlled, with the emergence of geometrically necessary dislocations and dislocation pile-ups playing a key role, and is strongly size dependent; and (iii) for wires with sufficiently large diameters plastic deformation becomes less heterogeneous and the dependence on size is greatly diminished., Comment: Accepted in Journal of the Mechanics and Physics of Solids

This paper is focused on the segregation of FGK dwarf and giant stars through narrow-band photometric data using the Spanish Virtual Observatory (SVO) Filter Profile Service and associated photometric tools. We selected spectra from the MILES, STELIB, and ELODIE stellar libraries, and used SVO photometric tools to derive the synthetic photometry in 15 J-PAS narrow filters, which were especially selected to cover spectral features sensitive to gravity changes. Using machine-learning techniques as the Gaussian mixture model and the support vector machine, we defined several criteria based on J-PAS colours to discriminate between dwarf and giant stars. We selected five colour-colour diagrams that presented the most promising separation between both samples. Our results show an overall accuracy in the studied sample of $\sim$0.97 for FGK stars, although a dependence on the luminosity type and the stellar effective temperature was found. We also defined a colour-temperature relation for dwarf stars with effective temperatures between 4\,000 and 7\,000\,K, which allows one to estimate the stellar effective temperature from four J-PAS filters ($J0450$, $J0510$, $J0550$, and $J0620$). Additionally, we extended the study to M-type giant and dwarf stars, achieving a similar accuracy to that for FGK stars., Comment: 19 pages, 18 figures, accepted for publication in A&A

We present Atacama Large Millimeter/submillimeter Array (ALMA) 1.3 mm continuum and CO(2-1) line emission observations toward the high-mass star formation region DR21. Five new continuum sources are found. We identify eighteen outflow streamers detected in CO emission radially arising from a common origin. The velocity spread of the outflow streamers range between $-$100 to $+$70 km s$^{-1}$. The radial velocities of each outflow roughly follow linear gradients (Hubble-Lemaitre-like expansion motions). Using the CO emission of the whole ensemble of streamers we estimate a total outflow mass of 120-210M$_{\odot}$. Additionally, we derived the dynamical age (8600 yr), momentum ($\sim10^{3}$ M$_{\odot}$ km s$^{-1}$), and kinetic energy ($\sim10^{48}$ erg) of the outflow. The morphology and kinematics presented by the CO outflow streamers confirm the presence of an explosive dispersal outflow at the heart of DR21. Five dispersal explosive outflows associated with massive star-forming regions have been confirmed in our Galaxy (Orion BN/KL, G5.89-0.39, S106-IR, IRAS16076-5134 and IRAS 12326-6245). However, their frequency of occurrence in the Galaxy and the originating nature are still uncertain., Comment: 19 pages, 15 figures including 4 in the appendix, accepted for publication to A&A

We report on the development of thermal detectors based on large-size tellurium dioxide crystals (45x45x45 mm), containing tellurium enriched in $^{130}$Te to about 91%, for the CROSS double-beta decay experiment. A powder used for the crystals growth was additionally purified by the directional solidification method, resulting in the reduction of the concentration of impurities by a factor 10, to a few ppm of the total concentration of residual elements (the main impurity is Fe). The purest part of the ingot (the first ~200 mm, about 80% of the total length of the cylindrical part of the ingot) was determined by scanning segregation profiles of impurities and used for the $^{130}$TeO$_2$ powder production with no evidence of re-contamination. The crystal growth was verified with precursors produced from powder with natural Te isotopic composition, and two small-size (20x20x10 mm) samples were tested at a sea-level laboratory showing high bolometric and spectrometric performance together with acceptable $^{210}$Po content (below 10 mBq/kg). This growth method was then applied for the production of six large cubic $^{130}$TeO$_2$ crystals and 4 of them were taken randomly to be characterized at the Canfranc underground laboratory, in the CROSS-dedicated low-background cryogenic facility. Two $^{130}$TeO$_2$ samples were coated with a thin, $O$(100 nm), metal film in form of Al layer (on 4 sides) or AlPd grid (on a single side) to investigate the possibility to tag surface events by pulse-shape discrimination. Similarly to the small natural precursors, large-volume $^{130}$TeO$_2$ bolometers show high performance and even better internal purity ($^{210}$Po activity $\sim$ 1 mBq/kg, while activities of $^{228}$Th and $^{226}$Ra are below 0.01 mBq/kg), satisfying requirements for the CROSS and, potentially, next-generation experiments., Comment: Submitted to JINST; 22 pages, 11 figures, 4 tables

The organizations and researchers producing research software face a common problem of making their software sustainable beyond funding provided by a single research project. This is addressed by research software engineers through building communities around their software, providing appropriate licensing, creating reliable and reproducible research software, making it sustainable and impactful, promoting, and ensuring that the research software is easy to adopt in research workflows, etc. As a result, numerous practices and guidelines exist to enhance research software quality, reusability, and sustainability. However, there is a lack of a unified framework to systematically integrate these practices and help organizations and research software developers refine their development and management processes. Our paper aims at bridging this gap by introducing a novel framework: RSMM. It is designed through systematic literature review and insights from interviews with research software project experts. In short, RSMM offers a structured pathway for evaluating and refining research software project management by categorizing 79 best practices into 17 capabilities across 4 focus areas. From assessing code quality and security to measuring impact, sustainability, and reproducibility, the model provides a complete evaluation of a research software project maturity. With RSMM, individuals as well as organizations involved in research software development gain a systematic approach to tackling various research software engineering challenges. By utilizing RSMM as a comprehensive checklist, organizations can systematically evaluate and refine their project management practices and organizational structure., Comment: 10 pages, 6 figures, 1 table. Submitted to IEEE eScience 2024 Conference

The CROSS experiment will search for neutrinoless double-beta decay using a specific mechanical structure to hold thermal detectors. The design of the structure was tuned to minimize the background contribution, keeping an optimal detector performance. A single module of the structure holds two scintillating bolometers (with a crystal size of 45x45x45 mm and a Ge slab facing the crystal's upper side) in the Cu frame, allowing for a modular construction of a large-scale array. Two designs are released: the initial $Thick$ version contains around 15% of Cu over the crystal mass (lithium molybdate, LMO), while this ratio is reduced to ~6% in a finer ($Slim$) design. Both designs were tested extensively at aboveground (IJCLab, France) and underground (LSC, Spain) laboratories. In particular, at LSC we used a pulse-tube-based CROSS facility to operate a 6-crystal array of LMOs enriched/depleted in $^{100}$Mo. The tested LMOs show high spectrometric performance in both designs; notably, the measured energy resolution is 5--7 keV FWHM at 2615 keV $\gamma$s, nearby the Q-value of $^{100}$Mo (3034 keV). Due to the absence of a reflective cavity around LMOs, a low scintillation signal is detected by Ge bolometers: ~0.3 keV (150 photons) for 1-MeV $\gamma$($\beta$) LMO-event. Despite that, an acceptable separation between $\alpha$ and $\gamma$($\beta$) events is achieved with most devices. The highest efficiency is reached with light detectors in the $Thick$ design thanks to a lower baseline noise width (0.05--0.09 keV RMS) when compared to that obtained in the $Slim$ version (0.10--0.35 keV RMS). Given the pivotal role of bolometric photodetectors for particle identification and random coincidences rejection, we will use the structure here described with upgraded light detectors, featuring thermal signal amplification via the Neganov-Trofimov-Luke effect, as also demonstrated in the present work., Comment: Submitted to JINST; 31 pages, 16 figures, 5 tables

The Artificial Intelligence Satellite Telecommunications Testbed (AISTT), part of the ESA project SPAICE, is focused on the transformation of the satellite payload by using artificial intelligence (AI) and machine learning (ML) methodologies over available commercial off-the-shelf (COTS) AI-capable chips for onboard processing. The objectives include validating artificial intelligence-driven SATCOM scenarios such as interference detection, spectrum sharing, radio resource management, decoding, and beamforming. The study highlights hardware selection and payload architecture. Preliminary results show that ML models significantly improve signal quality, spectral efficiency, and throughput compared to conventional payload. Moreover, the testbed aims to evaluate the performance and the use of AI-capable COTS chips in onboard SATCOM contexts., Comment: 6 pages, 4 figures

Despite the many advantages that ontology-based data access (OBDA) has brought to a range of application domains, state-of-the-art OBDA systems still do not support popular graph database management systems such as Neo4j. Algorithms for query rewriting focus on languages like conjunctive queries and their unions, which are fragments of first-order logic and were developed for relational data. Such query languages are poorly suited for querying graph data. Moreover, they also limit the expressiveness of the ontology languages that admit rewritings, restricting them to those where the data complexity of reasoning is not higher than it is in first-order logic. In this paper, we propose a technique for rewriting a family of navigational queries for a suitably restricted fragment of ELHI that extends DL-Lite and that is NL-complete in data complexity. We implemented a proof-of-concept prototype that rewrites into Cypher queries, and tested it on a real-world cognitive neuroscience use case with promising results., Comment: submitted to DL 2024 workshop

Centaurs, distinguished by their volatile-rich compositions, play a pivotal role in understanding the formation and evolution of the early solar system, as they represent remnants of the primordial material that populated the outer regions. Stellar occultations offer a means to investigate their physical properties, including shape, rotational state, or the potential presence of satellites and rings. This work aims to conduct a detailed study of the centaur (54598) Bienor through stellar occultations and rotational light curves from photometric data collected during recent years. We successfully predicted three stellar occultations by Bienor, which were observed from Japan, Eastern Europe, and the USA. In addition, we organized observational campaigns from Spain to obtain rotational light curves. At the same time, we develop software to generate synthetic light curves from three-dimensional shape models, enabling us to validate the outcomes through computer simulations. We resolve Bienor's projected ellipse for December 26, 2022, determine a prograde sense of rotation, and confirm an asymmetric rotational light curve. We also retrieve the axes of its triaxial ellipsoid shape as a = (127 $\pm$ 5) km, b = (55 $\pm$ 4) km, and c = (45 $\pm$ 4) km. Moreover, we refine the rotation period to 9.1736 $\pm$ 0.0002 hours and determine a geometric albedo of (6.5 $\pm$ 0.5) %, higher than previously determined by other methods. Finally, by comparing our findings with previous results and simulated rotational light curves, we analyze whether an irregular or contact-binary shape, the presence of an additional element such as a satellite, or significant albedo variations on Bienor's surface, may be present.

We compare the various chirality measures most widely used in the literature to quantify chiral symmetry in extended solids, i.e., the continuous chirality measure, the Hausdorff distance, and the angular momentum. By studying these functions in an algebraically tractable case, we can evaluate their strengths and weaknesses when applied to more complex crystals. Going beyond those classical calculations, we propose a new method to quantify the handedness of a crystal based on a pseudoscalar function, i.e., the helicity. This quantity, borrowed from hydrodynamics, can be computed from the eigenvector carrying the system from the high-symmetry non-chiral phase to the low-symmetry chiral phase. Different model systems like K$_3$NiO$_2$, CsCuCl$_3$ and MgTi$_2$O$_4$ are used as test cases where we show the superior interest of using helicity to quantify chirality together with the handedness distinction.

Semantic relations among entities are a widely accepted method for relation extraction. PromptORE (Prompt-based Open Relation Extraction) was designed to improve relation extraction with Large Language Models on generalistic documents. However, it is less effective when applied to historical documents, in languages other than English. In this study, we introduce an adaptation of PromptORE to extract relations from specialized documents, namely digital transcripts of trials from the Spanish Inquisition. Our approach involves fine-tuning transformer models with their pretraining objective on the data they will perform inference. We refer to this process as "biasing". Our Biased PromptORE addresses complex entity placements and genderism that occur in Spanish texts. We solve these issues by prompt engineering. We evaluate our method using Encoder-like models, corroborating our findings with experts' assessments. Additionally, we evaluate the performance using a binomial classification benchmark. Our results show a substantial improvement in accuracy -up to a 50% improvement with our Biased PromptORE models in comparison to the baseline models using standard PromptORE.

Recall that a defective $(m,n)$-parking function with defect $d$ is a parking function with $m$ cars attempting to park on a street with $n$ parking spots in which exactly $d$ cars fail to park. We establish a way to compute the defect of a defective $(m,n)$-parking function and show that the defect of a parking function is invariant under the action of $\mathfrak{S}_m$ the symmetric group on $[m]=\{1,2,\ldots,m\}$. We also show that the set of nondecreasing defective $(m,n)$-parking functions with defect $d$ are in bijection with the set of standard Young tableaux of shape $(n + d, m - d)$. This implies that the number of $\mathfrak{S}_m$-orbits of defective $(m,n)$-parking functions with defect $d$ is given by $\frac{n-m+2d+1}{n+d+1}\binom{m+n}{n+d}$. We also give a multinomial formula for the size of an $\mathfrak{S}_m$-orbit of a nondecreasing $(m,n)$-parking function with defect $d$. We conclude by using these results to give a new formula for the number of defective parking functions., Comment: 17 pages, 3 figures, 1 table

Here we present the discovery and characterization of the kagome metal TbTi$_3$Bi$_4$ in tandem with a new series of compounds, the Ln$_{2-x}$Ti$_{6+x}$Bi$_9$ (Ln: Tb-Lu) shurikagome metals. We previously reported on the growth of the LnTi$_3$Bi$_4$ (Ln: La-Gd$^{3+}$, Eu$^{2+}$, Yb$^{2+}$) family, a chemically diverse and exfoliable series of kagome metals with complex and highly anisotropic magnetism. However, unlike the La-Gd analogs, TbTi$_3$Bi$_4$ cannot be synthesized by our previous methodology due to phase competition with Ln$_{2-x}$Ti$_{6+x}$Bi$_9$ (x$\sim$1.7-1.2). Here we discuss the phase competition between the LnTi$_3$Bi$_4$ and Ln$_{2-x}$Ti$_{6+x}$Bi$_9$ families, helping to frame the difficulty in synthesizing LnTi$_3$Bi$_4$ compounds with small Ln species and providing a strategy to circumvent formation of Ln$_{2-x}$Ti$_{6+x}$Bi$_9$. Detailed characterization of the magnetic and electronic transport properties on single crystals of TbTi$_3$Bi$_4$ reveals a highly complex landscape of magnetic phases arising from an antiferromagnetic ground state. A series of metamagnetic transitions creates at least 5 unique magnetic phase pockets, including a 1/3 and 2/3 magnetization plateau. Further, the system exhibits an intimate connection between the magnetism and magnetotransport, exhibiting sharp switching from positive (+40%) to negative magnetoresistance (-50%). Like the LnTi$_3$Bi$_4$ kagome metals, the Ln$_{2-x}$Ti$_{6+x}$Bi$_9$ family exhibits quasi-2D networks of titanium and chains of rare-earth. We present the structures and some basic magnetic properties of the Ln$_{2-x}$Ti$_{6+x}$Bi$_9$ family alongside our characterization of the newly discovered TbTi$_3$Bi$_4$.

The sphere graph of $M_r$, a connect sum of $r$ copies of $S^1\times S^2$ was introduced by Hatcher as an analog of the curve graph of a surface to study the outer automorphism group of a free group $F_r$. Bestvina, Bromberg, and Fujiwara proved that the chromatic number of the curve graph is finite; bounds were subsequently improved by Gaster, Greene, and Vlamis. Motivated by the analogy, we provide upper and lower bounds for the chromatic number of the sphere graph of $M_r$. As a corollary to the prime decomposition of 3-manifolds, this gives bounds on the chromatic number of the sphere graph for any orientable 3-manifold., Comment: 7 pages, 2 figures

Electrical readout of magnetic states is a key to realize novel spintronics devices for efficient computing and data storage. Unidirectional magnetoresistance (UMR) in bilayer systems, consisting of a spin source material and a magnetic layer, refers to a change in the longitudinal resistance upon the reversal of magnetization, which typically originates from the interaction of spin-current and magnetization at the interface. Because of UMR s linear dependence on applied charge current and magnetization, it can be used to electrically read the magnetization state. However, in conventional spin source materials, the spin polarization of an electric field induced spin current is restricted to be in the film plane and hence the ensuing UMR can only respond to the in plane component of the magnetization. On the other hand, magnets with perpendicular magnetic anisotropy (PMA) are highly desired for magnetic memory and spin-logic devices, while the electrical read out of PMA magnets through UMR is critically missing. Here, we report the discovery of an unconventional UMR in bilayer heterostructures of a topological semimetal (WTe2) and a PMA ferromagnetic insulator (Cr2Ge2Te6, CGT), which allows to electrically read the up and down magnetic states of the CGT layer by measuring the longitudinal resistance. Our theoretical calculations based on a tight binding model show that the unconventional UMR originates from the interplay of crystal symmetry breaking in WTe2 and magnetic exchange interaction across the WTe2 and CGT interface. Combining with the ability of WTe2 to obtain magnetic field free switching of the PMA magnets, our discoveries open an exciting pathway to achieve two terminal magnetic memory devices that operate solely on the spin orbit torque and UMR, which is critical for developing next-generation non volatile and low power consumption data storage technologies.

Low-Rank Adaptation (LoRA) is a widely-used parameter-efficient finetuning method for large language models. LoRA saves memory by training only low rank perturbations to selected weight matrices. In this work, we compare the performance of LoRA and full finetuning on two target domains, programming and mathematics. We consider both the instruction finetuning (approximately 100K prompt-response pairs) and continued pretraining (20B unstructured tokens) data regimes. Our results show that, in the standard low-rank settings, LoRA substantially underperforms full finetuning. Nevertheless, LoRA better maintains the base model's performance on tasks outside the target domain. We show that LoRA mitigates forgetting more than common regularization techniques such as weight decay and dropout; it also helps maintain more diverse generations. Finally, we show that full finetuning learns perturbations with a rank that is 10-100X greater than typical LoRA configurations, possibly explaining some of the reported gaps. We conclude by proposing best practices for finetuning with LoRA., Comment: Final version with new experiments and analyses, as accepted to Transactions on Machine Learning Research, August 2024 (Featured Certification). https://openreview.net/forum?id=aloEru2qCG&noteId=Jb3PQNQDI2

We introduce a purely geometric formulation for two different measures addressed to quantify the entanglement between different parts of a tripartite qubit system. Our approach considers the entanglement-polytope defined by the smallest eigenvalues of the reduced density matrices of the qubit-components. The measures identify global and genuine entanglement, and are respectively associated with the projection and rejection of a given point of the polytope on the corresponding biseparable segments. Solving the so called `inverse problem', we also discuss a way to force the system to behave in a particular form, which opens the possibility of controlling and manipulating entanglement for practical purposes., Comment: 22+5 pages, 11+2 figures, 1 table. We have added an appendix with detailed information about the classification of three-qubit states in standard form (it includes 3 new tables). Some additional references are also included

This paper primarily focuses on the practical applications of optimal control theory for perturbed sweeping processes within the realm of robotics dynamics. By describing these models as controlled sweeping processes with pointwise control and state constraints and by employing necessary optimality conditions for such systems, we formulate optimal control problems suitable to these models and develop numerical algorithms for their solving. Subsequently, we use the Python Dynamic Optimization library GEKKO to simulate solutions to the posed robotics problems in the case of any fixed number of robots under different initial conditions.

Increasing agricultural productivity is a gradual process with significant time lags between research and development (R&D) investment and the resulting gains. We estimate the response of US agricultural Total Factor Productivity (TFP) to both R&D investment and weather, and quantify the public R&D spending required to offset the emerging impacts of climate change. We find that offsetting the climate-induced productivity slowdown by 2050 alone requires a sustained public R&D spending growth of 5.2-7.8% per year over 2021-2050. This amounts to an additional $208-$434B investment over this period. These are substantial requirements comparable to the public R&D spending growth that followed the two World Wars., Comment: Main text: 19 pages, 4 figures. Supplementary material: 47 pages, 20 figures, 13 tables

Dusty star-forming galaxies (DSFGs) contribute significantly to the stellar buildup at cosmic noon. Major mergers and gas accretion are often invoked to explain DSFGs' prodigious star-formation rates (SFRs) and large stellar masses. We conducted a spatially-resolved morphological analysis of the rest-frame UV/NIR emission in three DSFGs at z~2.5. Initially discovered as CO emitters by NOEMA observations of a bright Herschel source, we observed them with the JWST/NIRCam as part of the PEARLS program. The NIRCam data reveal the galaxies' stellar populations and dust distributions on scales of 250 pc. Spatial variations in stellar mass, SFR, and dust extinction are determined in resolved maps obtained through pixel-based SED fitting. The CO emitters are massive, dusty starburst galaxies with SFRs=340-2500 Msun/yr, positioning them among the most active SFGs at 2

Model merging and task arithmetic have emerged as promising scalable approaches to merge multiple single-task checkpoints to one multi-task model, but their applicability is reduced by significant performance loss. Previous works have linked these drops to interference in the weight space and erasure of important task-specific features. Instead, in this work we show that the information required to solve each task is still preserved after merging as different tasks mostly use non-overlapping sets of weights. We propose TALL-masks, a method to identify these task supports given a collection of task vectors and show that one can retrieve >99% of the single task accuracy by applying our masks to the multi-task vector, effectively compressing the individual checkpoints. We study the statistics of intersections among constructed masks and reveal the existence of selfish and catastrophic weights, i.e., parameters that are important exclusively to one task and irrelevant to all tasks but detrimental to multi-task fusion. For this reason, we propose Consensus Merging, an algorithm that eliminates such weights and improves the general performance of existing model merging approaches. Our experiments in vision and NLP benchmarks with up to 20 tasks, show that Consensus Merging consistently improves existing approaches. Furthermore, our proposed compression scheme reduces storage from 57Gb to 8.2Gb while retaining 99.7% of original performance., Comment: Accepted ICML 2024; The first two authors contributed equally to this work; Project website: https://tall-masks.github.io

Transformers are crucial for reliable and efficient power system operations, particularly in supporting the integration of renewable energy. Effective monitoring of transformer health is critical to maintain grid stability and performance. Thermal insulation ageing is a key transformer failure mode, which is generally tracked by monitoring the hotspot temperature (HST). However, HST measurement is complex, costly, and often estimated from indirect measurements. Existing HST models focus on space-agnostic thermal models, providing worst-case HST estimates. This article introduces a spatio-temporal model for transformer winding temperature and ageing estimation, which leverages physics-based partial differential equations (PDEs) with data-driven Neural Networks (NN) in a Physics Informed Neural Networks (PINNs) configuration to improve prediction accuracy and acquire spatio-temporal resolution. The computational accuracy of the PINN model is improved through the implementation of the Residual-Based Attention (PINN-RBA) scheme that accelerates the PINN model convergence. The PINN-RBA model is benchmarked against self-adaptive attention schemes and classical vanilla PINN configurations. For the first time, PINN based oil temperature predictions are used to estimate spatio-temporal transformer winding temperature values, validated through PDE numerical solution and fiber optic sensor measurements. Furthermore, the spatio-temporal transformer ageing model is inferred, which supports transformer health management decision-making. Results are validated with a distribution transformer operating on a floating photovoltaic power plant., Comment: 23 pages, 18 figures

Large language models (LLMs) are becoming bigger to boost performance. However, little is known about how explainability is affected by this trend. This work explores LIME explanations for DeBERTaV3 models of four different sizes on natural language inference (NLI) and zero-shot classification (ZSC) tasks. We evaluate the explanations based on their faithfulness to the models' internal decision processes and their plausibility, i.e. their agreement with human explanations. The key finding is that increased model size does not correlate with plausibility despite improved model performance, suggesting a misalignment between the LIME explanations and the models' internal processes as model size increases. Our results further suggest limitations regarding faithfulness metrics in NLI contexts., Comment: Published at ICLR 2024 Workshop on Secure and Trustworthy Large Language Models

This research examines the effectiveness of OpenAI's GPT models as independent evaluators of text summaries generated by six transformer-based models from Hugging Face: DistilBART, BERT, ProphetNet, T5, BART, and PEGASUS. We evaluated these summaries based on essential properties of high-quality summary - conciseness, relevance, coherence, and readability - using traditional metrics such as ROUGE and Latent Semantic Analysis (LSA). Uniquely, we also employed GPT not as a summarizer but as an evaluator, allowing it to independently assess summary quality without predefined metrics. Our analysis revealed significant correlations between GPT evaluations and traditional metrics, particularly in assessing relevance and coherence. The results demonstrate GPT's potential as a robust tool for evaluating text summaries, offering insights that complement established metrics and providing a basis for comparative analysis of transformer-based models in natural language processing tasks., Comment: 10 pages, 5 figures

In this research, we uses the DistilBERT model to generate extractive summary and the T5 model to generate abstractive summaries. Also, we generate hybrid summaries by combining both DistilBERT and T5 models. Central to our research is the implementation of GPT-based refining process to minimize the common problem of hallucinations that happens in AI-generated summaries. We evaluate unrefined summaries and, after refining, we also assess refined summaries using a range of traditional and novel metrics, demonstrating marked improvements in the accuracy and reliability of the summaries. Results highlight significant improvements in reducing hallucinatory content, thereby increasing the factual integrity of the summaries., Comment: 9 pages, 3 figures

We consider a set of challenging sequential manipulation puzzles, where an agent has to interact with multiple movable objects and navigate narrow passages. Such settings are notoriously difficult for Task-and-Motion Planners, as they require interdependent regrasps and solving hard motion planning problems. In this paper, we propose to search over sequences of easier pick-and-place subproblems, which can lead to the solution of the manipulation puzzle. Our method combines a heuristic-driven forward search of subproblems with an optimization-based Task-and-Motion Planning solver. To guide the search, we introduce heuristics to generate and prioritize useful subgoals. We evaluate our approach on various manually designed and automatically generated scenes, demonstrating the benefits of auxiliary subproblems in sequential manipulation planning., Comment: Accepted to ICRA 2024

Satellite constellation systems are becoming more attractive to provide communication services worldwide, especially in areas without network connectivity. While optimizing satellite gateway placement is crucial for operators to minimize deployment and operating costs, reducing the number of gateways may require more inter-satellite link hops to reach the ground network, thereby increasing latency. Therefore, it is of significant importance to develop a framework that optimizes gateway placement, dynamic routing, and flow management in inter-satellite links to enhance network performance. To this end, we model an optimization problem as a mixed-integer problem with a cost function combining the number of gateways, flow allocation, and traffic latency, allowing satellite operators to set priorities based on their policies. Our simulation results indicate that the proposed approach effectively reduces the number of active gateways by selecting their most appropriate locations while balancing the trade-off between the number of gateways and traffic latency. Furthermore, we demonstrate the impact of different weights in the cost function on performance through comparative analysis.

Experiments aimed at direct searches for WIMP dark matter require highly effective reduction of backgrounds and control of any residual radioactive contamination. In particular, neutrons interacting with atomic nuclei represent an important class of backgrounds due to the expected similarity of a WIMP-nucleon interaction, so that such experiments often feature a dedicated neutron detector surrounding the active target volume. In the context of the development of DarkSide-20k detector at INFN Gran Sasso National Laboratory (LNGS), several R&D projects were conceived and developed for the creation of a new hybrid material rich in both hydrogen and gadolinium nuclei to be employed as an essential element of the neutron detector. Thanks to its very high cross-section for neutron capture, gadolinium is one of the most widely used elements in neutron detectors, while the hydrogen-rich material is instrumental in efficiently moderating the neutrons. In this paper results from one of the R&Ds are presented. In this effort the new hybrid material was obtained as a poly(methyl methacrylate) (PMMA) matrix, loaded with gadolinium oxide in the form of nanoparticles. We describe its realization, including all phases of design, purification, construction, characterization, and determination of mechanical properties of the new material.

Weakly interacting massive particles (WIMPs) are well-motivated candidates for dark matter. One signature of galactic WIMPs is the annual modulation expected in a detector's interaction rate, which arises from Earth's revolution around the Sun. Over two decades, the DAMA/LIBRA experiment has observed such modulation with 250 kg of NaI(Tl) scintillators, in accordance with WIMP expectations but inconsistent with the negative results of other experiments. The signal depends on the target material, so to validate or refute the DAMA result, the experiment must be replicated using the same material. This is the goal of the ANAIS-112 experiment, currently underway since August 2017 with 112.5 kg of NaI(Tl). In this work, we present a reanalysis of three years of data employing an improved analysis chain to enhance the experimental sensitivity. The results presented here are consistent with the absence of modulation and inconsistent with DAMA's observation at nearly 3$\sigma$ confidence level, with the potential to reach a 5$\sigma$ level within 8 years from the beginning of the data collection. Additionally, we explore the impact of different scintillation quenching factors in the comparison between ANAIS-112 and DAMA/LIBRA., Comment: 16 pages, 10 figures, 1 table

LP~349$-$25 is a well studied close stellar binary system comprised of two late M dwarf stars, both stars close to the limit between star and brown dwarf. This system was previously identified as a source of GHz radio emission. We observed LP~349$-$25AB in 11 epochs in 2020$-$2022, detecting both components in this nearby binary system using the Very Long Baseline Array (VLBA). We fit simultaneously the VLBA absolute astrometric positions together with existing relative astrometric observations derived from optical/infrared observations with a set of algorithms that use non-linear least-square, Genetic Algorithm and Markov Chain Montecarlo methods to determine the orbital parameters of the two components. We find the masses of the primary and secondary components to be 0.08188 $\pm$ 0.00061 $M\odot$, and 0.06411 $\pm$ 0.00049 $M_\odot$, respectively, representing one of the most precise mass estimates of any UCDs to date. The primary is an UCD of 85.71$\pm$0.64 M$_{Jup}$, while the secondary has a mass consistent with being a Brown Dwarf of 67.11 $\pm$ 0.51 $M_{Jup}$. This is one of the very few direct detections of a Brown Dwarf with VLBA observations. We also find a distance to the binary system of 14.122 $\pm$ 0.057 pc. Using Stellar Evolutionary Models, we find the model-derived stellar parameters of both stars. In particular, we obtain a model-derived age of 262 Myr for the system, which indicates that LP~349$-$25AB is composed of two pre-main-sequence stars. In addition, we find that the secondary star is significantly less evolved that the primary star.

We present results from a search for radio emission in 77 stellar systems hosting 140 exoplanets, predominantly within 17.5 pc using the Very Large Array (VLA) at $4-8$ GHz. This is the largest and most sensitive search to date for radio emission in exoplanetary systems in the GHz frequency range. We obtained new observations of 58 systems, and analyzed archival observations of an additional 19 systems. Our choice of frequency and volume limit are motivated by radio detections of ultracool dwarfs (UCDs), including T dwarfs with masses at the exoplanet threshold of $\sim\!13\,M_J$. Our surveyed exoplanets span a mass range of $\approx\,10^{-3}-10\,M_J$ and semi-major axes of $\approx\,10^{-2}-10\,$AU. We detect a single target - GJ 3323 (M4) hosting two exoplanets with minimum masses of 2 and 2.3$\,M_\oplus$ - with a circular polarization fraction of $\approx\,40\%$; the radio luminosity agrees with its known X-ray luminosity and the G\"udel-Benz relation for stellar activity suggesting a likely stellar origin, but the high circular polarization fraction may also be indicative of star-planet interaction. For the remaining sources our $3\sigma$ upper limits are generally $L_\nu\lesssim\,10^{12.5}\,\mathrm{erg}\,\mathrm{s}^{-1}\,\mathrm{Hz}^{-1}$, comparable to the lowest radio luminosities in UCDs. Our results are consistent with previous targeted searches of individual systems at GHz frequencies while greatly expanding the sample size. Our sensitivity is comparable to predicted fluxes for some systems considered candidates for detectable star-planet interaction. Observations with future instruments such as the Square Kilometer Array and Next Generation Very Large Array will be necessary to further constrain emission mechanisms from exoplanet systems at GHz frequencies., Comment: Submitted to ApJ, 18 pages, 8 figures

The main purpose of this paper is to study both the underdamped and the overdamped dynamics of the nonlinear Helmholtz oscillator with a fractional order damping. For that purpose, we use the Grunwald-Letnikov fractional derivative algorithm in order to get the numerical simulations. Here, we investigate the effect of taking the fractional derivative in the dissipative term in function of the parameter a. Our main findings show that the trajectories can remain inside the well or can escape from it depending on a which plays the role of a control parameter. Besides, the parameter a is also relevant for the creation or destruction of chaotic motions. On the other hand, the study of the escape times of the particles from the well, as a result of variations of the initial conditions and the undergoing force F, is reported by the use of visualization techniques such as basins of attraction and bifurcation diagrams, showing a good agreement with previous results. Finally, the study of the escape times versus the fractional parameter a shows an exponential decay which goes to zero when a is larger than one. All the results have been carried out for weak damping where chaotic motions can take place in the non-fractional case and also for a stronger damping (overdamped case), where the influence of the fractional term plays a crucial role to enhance chaotic motions. We expect that these results can be of interest in the field of fractional calculus and its applications.

Global communications have undergone a paradigm shift with the rapid expansion of low-earth orbit (LEO) satellite constellations, offering a new space era of reduced latency and ubiquitous, high-speed broadband internet access. However, the fast developments in LEO orbits pose significant challenges, particularly the coexistence with geostationary earth orbit (GEO) satellite systems. This article presents an overview of the regulatory aspects that cover the spectrum sharing in the bands allocated to the Fixed Satellite Service between geostationary networks (GSO) and non-geostationary systems (NGSO), as well as the main interference mitigation techniques for their coexistence. Our work highlights the increased potential for inter-system interference. It explores the regulatory landscape following the World Radio Conference (WRC-23). We discuss the different interference management strategies proposed for the GSO-NGSO spectral coexistence, including on-board and ground-based approaches and more advanced mitigation techniques based on beamforming. Moving onto operational aspects related to the sharing of spectrum, we introduce recent work on interference detection, identification, and mitigation and provide our vision of the emerging role of artificial intelligence (AI) in the aforementioned tasks., Comment: magazine under review

Diamond is increasingly popular because of its unique material properties. Diamond defects called nitrogen vacancy (NV) centers allow measurements with unprecedented sensitivity. However, to achieve ideal sensing performance NV centers need to be within nanometers from the surface and are thus strongly dependent on the local surface chemistry. Several attempts have been made to compare diamond surfaces. However, due to the high price of diamond crystals with shallow NV centers, a limited number of chemical modifications have been studied. Here, we developed a systematic method to investigate a continuity of different local environments with a varying density and nature of surface groups in a single experiment on a single diamond plate. To achieve this goal, we used diamonds with a shallow ensemble of NV centers and introduced a chemical gradient across the surface. More specifically we used air and hydrogen plasma. The gradients were formed by low pressure plasma treatment after masking with a right-angled triangular prism shield. As a result, the surface contained gradually more oxygen/hydrogen towards the open end of the shield. We then performed widefield relaxometry to determine the effect of surface chemistry on the sensing performance. As expected, relaxation times and thus sensing performance indeed varies along the gradient., Comment: Main text: 18 pages and 9 figures Supplementary materials: 6 pages and 6 figures

We present an analysis of rest-frame UV continuum slopes, $\beta$, using a sample of 1011 galaxies at $6.5

The first public 0.9-4.4{\mu}m NIRCam images of the North Ecliptic Pole (NEP) Time Domain Field (TDF) uncovered galaxies displaying point-source features in their cores as seen in the longer wavelength filters. We visually identified a sample of 66 galaxies (~1 galaxy per arcmin2) with point-like cores and have modeled their two-dimensional light profiles with GalFit, identifying 16 galactic nuclei with measurable point-source components. GalFit suggests the visual sample is a mix of both compact stellar bulge and point-source galaxy cores. This core classification is complemented by spectral energy distribution (SED) modeling to infer the sample's active galactic nucleus (AGN) and host-galaxy parameters. For galaxies with measurable point-source components, the median fractional AGN contribution to their 0.1-30.0{\mu}m flux is 0.44, and 14/16 are color-classified AGN. We conclude that near-infrared point-source galaxy cores are signatures of AGN. In addition, we define an automated sample-selection criterion to identify these point-source features. These criteria can be used in other extant and future NIRCam images to streamline the search for galaxies with unresolved IR-luminous AGN. The James Webb Space Telescope's superb angular resolution and sensitivity at infrared wavelengths is resurrecting the morphological identification of AGN., Comment: 17 pages, 9 figures, 2 tables

The large spin-orbit coupling in iridium oxides plays a significant role in driving novel physical behaviors, including emergent phenomena in the films and heterostructures of perovskite and Ruddlesden-Popper iridates. In this work, we study the role of epitaxial strain on the electronic behavior of thin SrIrO$_3$ films. We find that compressive epitaxial strain leads to metallic transport behavior, but a slight tensile strain shows gapped behavior. Temperature-dependent resistivity measurements are used to examine different behaviors in films as a function of strain. We find Kondo contributions to the resistivity, with stronger effects in films that are thinner and under less compressive epitaxial strain. These results show the potential to tune SrIrO$_3$ into Kondo insulating states and open possibilities for a quantum critical point that can be controlled with strain in epitaxial films., Comment: 13 pages, 4 figures