Search

Large Language Models (LLMs) have demonstrated significant potential in handling specialized tasks, including medical problem-solving. However, most studies predominantly focus on English-language contexts. This study introduces a novel benchmark dataset based on Polish medical licensing and specialization exams (LEK, LDEK, PES) taken by medical doctor candidates and practicing doctors pursuing specialization. The dataset was web-scraped from publicly available resources provided by the Medical Examination Center and the Chief Medical Chamber. It comprises over 24,000 exam questions, including a subset of parallel Polish-English corpora, where the English portion was professionally translated by the examination center for foreign candidates. By creating a structured benchmark from these existing exam questions, we systematically evaluate state-of-the-art LLMs, including general-purpose, domain-specific, and Polish-specific models, and compare their performance against human medical students. Our analysis reveals that while models like GPT-4o achieve near-human performance, significant challenges persist in cross-lingual translation and domain-specific understanding. These findings underscore disparities in model performance across languages and medical specialties, highlighting the limitations and ethical considerations of deploying LLMs in clinical practice.

Space debris presents a critical challenge for the sustainability of future space missions, emphasizing the need for robust and standardized identification methods. However, a comprehensive benchmark for rocket body classification remains absent. This paper addresses this gap by introducing the RoBo6 dataset for rocket body classification based on light curves. The dataset, derived from the Mini Mega Tortora database, includes light curves for six rocket body classes: CZ-3B, Atlas 5 Centaur, Falcon 9, H-2A, Ariane 5, and Delta 4. With 5,676 training and 1,404 test samples, it addresses data inconsistencies using resampling, normalization, and filtering techniques. Several machine learning models were evaluated, including CNN and transformer-based approaches, with Astroconformer reporting the best performance. The dataset establishes a common benchmark for future comparisons and advancements in rocket body classification tasks., Comment: 6 pages, 1 figure, 5 tables, Accepted on Machine Learning and the Physical Sciences Workshop, NeurIPS 2024

In this paper we provide the non-existence criterion for the so-called maximizing curves of odd degrees. Furthermore, in the light of our criterion, we define a new class of plane curves that generalizes the notion of maximizing curves which we call as $M$-curves., Comment: 10 pages

Every year, thousands of people suffer from skull damage and require personalized implants to fill the cranial cavity. Unfortunately, the waiting time for reconstruction surgery can extend to several weeks or even months, especially in less developed countries. One factor contributing to the extended waiting period is the intricate process of personalized implant modeling. Currently, the preparation of these implants by experienced biomechanical experts is both costly and time-consuming. Recent advances in artificial intelligence, especially in deep learning, offer promising potential for automating the process. However, deep learning-based cranial reconstruction faces several challenges: (i) the limited size of training datasets, (ii) the high resolution of the volumetric data, and (iii) significant data heterogeneity. In this work, we propose a novel approach to address these challenges by enhancing the reconstruction through learnable symmetry enforcement. We demonstrate that it is possible to train a neural network dedicated to calculating skull symmetry, which can be utilized either as an additional objective function during training or as a post-reconstruction objective during the refinement step. We quantitatively evaluate the proposed method using open SkullBreak and SkullFix datasets, and qualitatively using real clinical cases. The results indicate that the symmetry-preserving reconstruction network achieves considerably better outcomes compared to the baseline (0.94/0.94/1.31 vs 0.84/0.76/2.43 in terms of DSC, bDSC, and HD95). Moreover, the results are comparable to the best-performing methods while requiring significantly fewer computational resources (< 500 vs > 100,000 GPU hours). The proposed method is a considerable contribution to the field of applied artificial intelligence in medicine and is a step toward automatic cranial defect reconstruction in clinical practice.

We use time-resolved thermometry to monitor the decay of nonequilibrium quasiparticles in superconducting Al in the temperature range from $0.3\,$K to $1.2\,$K. The quasiparticle lifetime at higher temperatures ($T>0.7\,$K) agrees well with the calculated energy flow from electrons to phonons, but at lower temperatures it is significantly shorter than the theory predicts. We show well-defined internal equilibrium of quasiparticle system in the studied thermal transients, which implicates that quasiparticle-quasiparticle relaxation is much faster than electron-phonon interaction.

We initiate a systematic study of utilizing predictions to improve over approximation guarantees of classic algorithms, without increasing the running time. We propose a systematic method for a wide class of optimization problems that ask to select a feasible subset of input items of minimal (or maximal) total weight. This gives simple (near-)linear time algorithms for, e.g., Vertex Cover, Steiner Tree, Min-Weight Perfect Matching, Knapsack, and Clique. Our algorithms produce optimal solutions when provided with perfect predictions and their approximation ratios smoothly degrade with increasing prediction error. With small enough prediction error we achieve approximation guarantees that are beyond reach without predictions in the given time bounds, as exemplified by the NP-hardness and APX-hardness of many of the above problems. Although we show our approach to be optimal for this class of problems as a whole, there is a potential for exploiting specific structural properties of individual problems to obtain improved bounds; we demonstrate this on the Steiner Tree problem. We conclude with an empirical evaluation of our approach.

We investigate the hydrodynamic solutions for expanding bubbles in cosmological first-order phase transitions going beyond local thermal equilibrium approximation. Under the assumption of a tangenosidal field profile, we supplement the matching conditions with the entropy produced due to the interaction of the bubble wall with ambient plasma. This allows us to analytically compute the corresponding fluid profiles and find bubble-wall velocity. We show that due to the entropy production, two stable solutions corresponding to a deflagration or hybrid and a detonation can coexist. Finally, we use numerical real-time simulations of bubble growth to show that in such cases it is typically the faster detonation solution which is realised. This effect can be explained in terms of the fluid profile not being fully formed into the predicted steady-state solution as the wall accelerates past this slower solution., Comment: 18 pages, 4 figures

Fiber-based long-haul quantum communication would greatly benefit from a robust and deterministically integrated source of quantum state. Here, we report the design, fabrication, and optical characterization of InAs/InP quantum dots in the InP H1 point defect 2D photonic crystal cavity, integrated with the standard single-mode fiber using a micro-transfer printing technique. The device was placed in a compact cryocooler maintaining a cryogenic temperature of 15 K with single-photon emission characterized by $g^{(2)}(0)=0.14(14)$ and reliable and stable emission (intensity fluctuations given by a standard deviation $\sigma = 0.13$), so that an all-fiber based connection between two laboratory nodes through an open area was established and utilized for testing the quantum channel. In this way, we demonstrate a plug-and-play all-fiber single-photon source operating in the third telecom window, where standard telecommunication fiber networks can be used as a low-loss medium.

Optimization problems pose challenges across various fields. In recent years, quantum annealers have emerged as a promising platform for tackling such challenges. To provide a new perspective, we develop a heuristic tensor-network-based algorithm to reveal the low-energy spectrum of Ising spin-glass systems with interaction graphs relevant to present-day quantum annealers. Our deterministic approach combines a branch-and-bound search strategy with an approximate calculation of marginals via tensor-network contractions. Its application to quasi-two-dimensional lattices with large unit cells of up to 24 spins, realized in current quantum annealing processors, requires a dedicated approach that utilizes sparse structures in the tensor network representation and GPU hardware acceleration. We benchmark our approach on random problems defined on Pegasus and Zephyr graphs with up to a few thousand spins, comparing it against the D-Wave Advantage quantum annealer and Simulated Bifurcation algorithm, with the latter representing an emerging class of classical Ising solvers. Apart from the quality of the best solutions, we compare the diversity of low-energy states sampled by all the solvers. For the biggest considered problems with over 5000 spins, the state-of-the-art tensor network approaches lead to solutions that are $0.1\%$ to $1\%$ worse than the best solutions obtained by Ising machines while being two orders of magnitude slower. We attribute those results to approximate contraction failures. While all three methods can output diverse low-energy solutions, e.g., differing by at least a quarter of spins with energy error below $1\%$, our deterministic branch-and-bound approach finds sets of a few such states at most. On the other hand, both Ising machines prove capable of sampling sets of thousands of such solutions., Comment: 15+7 pages, 12+7 figures

We study the evolution of growing vacuum bubbles. The bubble walls interact with the surrounding fluid and may, consequently, reach a terminal velocity. If the mean free path of the particles in the fluid is much shorter than the bubble wall thickness, the fluid is locally in thermal equilibrium and the wall's terminal velocity can be determined by entropy conservation. On the other hand, if local thermal equilibrium inside the wall cannot be maintained, the wall velocity can be estimated from the pressure impacted by ballistic particle dynamics at the wall. We find that the latter case leads to slightly slower bubble walls. Expectedly, we find the largest differences in the terminal velocity when the fluid is entirely ballistic. This observation indicates that the non-equilibrium effects inside walls are relevant. To study bubble evolution, we perform hydrodynamic lattice simulations in the case of local thermal equilibrium and $N$-body simulations in the ballistic case to investigate the dynamical effects during expansion. Both simulations show that even if a stationary solution exists in theory it may not be reached depending on the dynamics of the accelerating bubble walls., Comment: 26 pages, 7 figures

A finite element model consisting of a conducting filament with or without a gap was used to reproduce behavior of TaO$_x$-based resistive switching devices. The specific goal was to explore the range of possible filament parameters such a filament diameter, composition, gap width, and composition to reproduce the conductance and shape of I-V while keeping the maximum temperature within acceptable range allowing for ion motion and preventing melting. The model solving heat and charge transport produced a good agreement with experimental data for the oxygen content in the filament below TaO$_{1.3}$, the filament diameter range between 6 and 22 nm, and the gap oxygen content between TaO$_{1.7}$ and TaO$_{1.85}$. Gap width was not limited on either low or high sides by the criteria considered in this report. The obtained filament composition corresponds to oxygen deficiency an order of magnitude higher than one estimated by other modelling efforts. This was in a large part due to use of recent experimental values of conductivity as a function of composition and temperature. Our modelling results imply that a large fraction of atoms leaves and/or accumulates within the filament to produce large relative concentration change. This, in turn, necessitates inclusion of strain energy in the filament formation modelling. In addition, the results reproduce non-linear I-V without the necessity of assuming the Poole-Frenkel type of electrical conduction or presence of a barrier at the oxide/metal interface., Comment: 31 pages, 7 figures

Though large language models (LLMs) have demonstrated exceptional performance across numerous problems, their application to predictive tasks in relational databases remains largely unexplored. In this work, we address the notion that LLMs cannot yield satisfactory results on relational databases due to their interconnected tables, complex relationships, and heterogeneous data types. Using the recently introduced RelBench benchmark, we demonstrate that even a straightforward application of LLMs achieves competitive performance on these tasks. These findings establish LLMs as a promising new baseline for ML on relational databases and encourage further research in this direction.

Model-based reinforcement learning refers to a set of approaches capable of sample-efficient decision making, which create an explicit model of the environment. This model can subsequently be used for learning optimal policies. In this paper, we propose a temporal Gaussian Mixture Model composed of a perception model and a transition model. The perception model extracts discrete (latent) states from continuous observations using a variational Gaussian mixture likelihood. Importantly, our model constantly monitors the collected data searching for new Gaussian components, i.e., the perception model performs a form of structure learning (Smith et al., 2020; Friston et al., 2018; Neacsu et al., 2022) as it learns the number of Gaussian components in the mixture. Additionally, the transition model learns the temporal transition between consecutive time steps by taking advantage of the Dirichlet-categorical conjugacy. Both the perception and transition models are able to forget part of the data points, while integrating the information they provide within the prior, which ensure fast variational inference. Finally, decision making is performed with a variant of Q-learning which is able to learn Q-values from beliefs over states. Empirically, we have demonstrated the model's ability to learn the structure of several mazes: the model discovered the number of states and the transition probabilities between these states. Moreover, using its learned Q-values, the agent was able to successfully navigate from the starting position to the maze's exit.

A computational study of I-BODIPY (2-ethyl-4,4-difluoro-6,7-diiodo-1,3-dimethyl-4-bora-3a,4a-diaza-s-indacene) was conducted to investigate its photophysical properties as a potential triplet photosensitizer for singlet oxygen generation. Multireference CASPT2 and CASSCF methods were used to calculate vertical excitation energies and spin-orbit couplings (SOCs) in a model monoiodinated BODIPY molecule to assess the applicability of the single-reference ADC(2) method. Time-dependent density functional theory (TD-DFT) with the Tamm-Dancoff approximation (TDA) was tested against ADC(2) using different exchange-correlation functionals, employing a two-component pseudopotential basis set for iodine. SOC magnitudes between excited states were discussed using the Slater-Condon rules. The geometry dependence of SOCs for the lowest states was also examined. TD-DFT/B3LYP and TD-DFT(TDA)/BHLYP were selected for subsequent absorption spectra and trajectory surface hopping (TSH) molecular dynamics (MD) simulations. Two bright states were identified in I-BODIPY's visible spectrum, showing a red shift due to iodine substitution. Excited-state MD simulations, including nonadiabatic effects and SOCs, were performed to investigate relaxation after photoexcitation to the S1 state. TSH MD simulations revealed that intersystem crossings occur on a similar timescale to internal conversions. After triplet population growth, a "saturation" phase was reached with a triplet-to-singlet ratio of about 4:1. The calculated triplet quantum yield of 0.85 agrees qualitatively with the experimental singlet oxygen generation yield of 0.99.

We examine the onset of Anderson localization in three-dimensional systems with structural disorder in the form of lattice irregularities and in the absence of any on-site disordered potential. Analyzing two models with distinct types of lattice regularities, we show that the Anderson localization transition occurs when the strength of the structural disorder is smoothly increased. Performing finite-size scaling analysis of the results, we show that the transition belongs to the same universality class as regular Anderson localization induced by onsite disorder. Our work identifies a new class of structurally disordered lattice models in which destructive interference of matter waves may inhibit transport and lead to a transition between metallic and localized phases., Comment: 8 pages, 6 figures

Nonlinear squeezing is a property of non-Gaussian states of light with an important application in continuous variable quantum computing. We study the generation of nonlinear squeezing in multimode systems produced by the photon-added coherent state technique. We present a protocol and find a regime in which the nonlinear squeezing appears in two modes simultaneously, even for a weak non-Gaussianity induced by the single-photon addition. We explore the properties of nonlinear squeezing depending on the modal structure of light, as well as the seed and local oscillator profiles, and present an optimal measurement strategy., Comment: 8 pages, 6 figures

Topology optimization is an essential tool in computational engineering, for example, to improve the design and efficiency of flow channels. At the same time, Ising machines, including digital or quantum annealers, have been used as efficient solvers for combinatorial optimization problems. Beyond combinatorial optimization, recent works have demonstrated applicability to other engineering tasks by tailoring corresponding problem formulations. In this study, we present a novel Ising machine formulation for computing design updates during topology optimization with the goal of minimizing dissipation energy in flow channels. We explore the potential of this approach to improve the efficiency and performance of the optimization process. To this end, we conduct experiments to study the impact of various factors within the novel formulation. Additionally, we compare it to a classical method using the number of optimization steps and the final values of the objective function as indicators of the time intensity of the optimization and the performance of the resulting designs, respectively. Our findings show that the proposed update strategy can accelerate the topology optimization process while producing comparable designs. However, it tends to be less exploratory, which may lead to lower performance of the designs. These results highlight the potential of incorporating Ising formulations for optimization tasks but also show their limitations when used to compute design updates in an iterative optimization process. In conclusion, this work provides an efficient alternative for design updates in topology optimization and enhances the understanding of integrating Ising machine formulations in engineering optimization.

Thermodynamics of local causal horizons have been shown to encode the information necessary to derive the equations governing the gravitational dynamics. We have previously shown that, in the presence of matter, this derivation further implies quantum phenomenological corrections to gravitational dynamics. Herein, we study whether similar corrections also occur in vacuum. We show that, under the assumptions of locality and local Lorentz invariance of physics, the vacuum dynamics is prescribed by the Einstein equations. We also discuss an alternative paradigm which assumes the existence of a preferred direction of time (much like in Einstein-aether or Ho\v{r}ava-Lifshitz gravity). Then, we find an modified gravitational dynamics in which Ricci curvature is sourced by Bel-Robinson super-energy., Comment: 16 pages, 1 figure

We consider the classical single-source shortest path problem in directed weighted graphs. D.~Eppstein proved recently an $\Omega(n^3)$ lower bound for oblivious algorithms that use relaxation operations to update the tentative distances from the source vertex. We generalize this result by extending this $\Omega(n^3)$ lower bound to \emph{adaptive} algorithms that, in addition to relaxations, can perform queries involving some simple types of linear inequalities between edge weights and tentative distances. Our model captures as a special case the operations on tentative distances used by Dijkstra's algorithm.

Thin films and van der Waals heterostructures derived from two-dimensional solids offer enormous potential for a broad range of novel, energy efficient devices, however, their use is currently hampered by slow, labor-intensive fabrication methods often employing hazardous chemicals. We demonstrate a novel technique for rapid, low-cost and environmentally-friendly production of ultra-thin films and van der Waals heterostructures of two-dimensional solids from aqueous surfactant-stabilized suspensions, which we term `Liquid Interface Deposition'. Films are produced by the transfer of platelets of two-dimensional materials assembled at the interface between two immiscible liquids (water and dicholoromethane), at which surfactants which stabilize the two-dimensional materials in the aqueous phase are `stripped' from the platelets due to preferential partitioning of the surfactant to the non-aqeuous phase. The approach is generic to two-dimensional materials which can be stabilized in aqueous suspension by a surfactant and the resulting films can then be transferred to an arbitrary substrate by a range of approaches. The generic applicability of this technique is demonstrated through production of thin films on a variety of substrates, deposition of transparent, highly conductive, graphene films with conductivities between $7.7 \times 10^{3} - 1.26\times 10^{5}~\textrm{S m}^{-1}$ and transmittances of 55-75%, and by the fabrication of van der Waals heterostructures of MoS$_2$, WS$_2$, and few-layer graphene.

External tools help large language models (LLMs) succeed at tasks where they would otherwise typically fail. In existing frameworks, LLMs learn tool use either by in-context demonstrations or via full model fine-tuning on annotated data. As these approaches do not easily scale, a recent trend is to abandon them in favor of lightweight, parameter-efficient tuning paradigms. These methods allow quickly alternating between the frozen LLM and its specialised fine-tuned version, by switching on or off a handful of additional custom parameters. Hence, we postulate that the generalization ability of the frozen model can be leveraged to improve tool selection. We present Tool selECTion via meta-reasONing (TECTON), a two-phase system that first reasons over a task using a custom fine-tuned LM head and outputs candidate tools. Then, with the custom head disabled, it meta-reasons (i.e., it reasons over the previous reasoning process) to make a final choice. We show that TECTON results in substantial gains - both in-distribution and out-of-distribution - on a range of math reasoning datasets.

The goal of this paper is to develop a multilingual classifier and conditional probability estimator of occupation codes for online job advertisements according in accordance with the International Standard Classification of Occupations (ISCO) extended with the Polish Classification of Occupations and Specializations (KZiS), which is analogous to the European Classification of Occupations. In this paper, we utilise a range of data sources, including a novel one, namely the Central Job Offers Database, which is a register of all vacancies submitted to Public Employment Offices. Their staff members code the vacancies according to the ISCO and KZiS. A hierarchical multi-class classifier has been developed based on the transformer architecture. The classifier begins by encoding the jobs found in advertisements to the widest 1-digit occupational group, and then narrows the assignment to a 6-digit occupation code. We show that incorporation of the hierarchical structure of occupations improves prediction accuracy by 1-2 percentage points, particularly for the hand-coded online job advertisements. Finally, a bilingual (Polish and English) and multilingual (24 languages) model is developed based on data translated using closed and open-source software. The open-source software is provided for the benefit of the official statistics community, with a particular focus on international comparability., Comment: 51 pages

Semiconducting transition metal dichalcogenides of group VI are well-known for their prominent excitonic effects and the transition from an indirect to a direct band gap when reduced to monolayers. While considerable efforts have elucidated the Rydberg series of excitons in monolayers, understanding their properties in multilayers remains incomplete. In these structures, despite an indirect band gap, momentum-direct excitons largely shape the optical response. In this work, we combine magneto-reflectance experiments with theoretical modeling based on the $\mathbf {k\cdot p}$ approach to investigate the origin of excitonic resonances in WSe$_2$ bi-, tri-, and quadlayers. For all investigated thicknesses, we observe a series of excitonic resonances in the reflectance spectra, initiated by a ground state with an amplitude comparable to the ground state of the 1$s$ exciton in the monolayer. Higher energy states exhibit a decrease in intensity with increasing energy, as expected for the excited states of the Rydberg series, although a significant increase in the diamagnetic shift is missing in tri- and quadlayers. By comparing the experimental observations with theoretical predictions, we discover that the excitonic resonances observed in trilayers originate from two Rydberg series, while quadlayers exhibit four such series, and bilayers host a single Rydberg series similar to that found in monolayers., Comment: 31 pages, 16 figures, published in Phys. Rev. B

This paper investigates Nash equilibria in pure strategies for quantum approach to the Prisoner's Dilemma. The quantization process involves extending the classical game by introducing two additional unitary strategies. We consider five classes of such quantum games, which remain invariant under isomorphic transformations of the classical game. For each class, we identify and analyse all possible Nash equilibria. Our results reveal the complexity and diversity of strategic behaviour in the quantum setting, providing new insights into the dynamics of classical decision-making dilemmas. In the case of the standard Prisoner's Dilemma, the resulting Nash equilibria of quantum extensions are found to be closer to Pareto optimal solutions than those of the classical equilibrium.

This work examines the hypothesis that intravoxel incoherent motion MRI (IVIM) can quantify local cerebral blood flow (qCBF), infarct volume, and define the ischemic penumbra for determination of the perfusion-diffusion mismatch (PWI/DWI) volume in a setting of acute ischemic stroke. Eight experiments were conducted in a pre-clinical middle cerebral artery occlusion (MCAO) model. IVIM and dynamic susceptibility contrast (DSC) imaging were acquired 2.5hr post-MCAO. IVIM was post-processed using software written in-house to produce parametric images of local qCBF, Water Transport Time (WTT), diffusion, and subsequently, PWI/DWI mismatch. These IVIM image parameters were compared with delay-and-dispersion-corrected local-AIF DSC perfusion image parameters including Tmax, qCBF, mean transit time (MTT), and mean diffusivity for DSC PWI/DWI mismatch. Final infarct volume was measured 4hrs post-occlusion. Early (2.5hr post-occlusion) DSC qCBF and IVIM qCBF in the diffusion negative MCA territory correlated strongly (slope=1.00, p=0.01,R2=0.69,Lins CCC=0.71), and both DSC and IVIM qCBF values negatively correlated with final infarct volume (R2=0.78,R2=0.61 respectively). The volume of hypoperfusion measured at 2.5 hours from DSC qCBF and from IVIM qCBF both predicted final infarct volume with good sensitivity and correlation (slope=2.08, R2=0.67, slope=2.50,R2=0.68 respectively). IVIM PWI/DWI ratio was correlated with infarct growth (R2=0.70) and WTT correlated with MTT (slope=0.82,R2=0.60). IVIM qCBF correlated strongly with local-AIF DSC qCBF and IVIM PWI/DWI correlated strongly with infarct growth. Both DSC and IVIM quantitative perfusion image acquired early after occlusion were able to predict final infarct volume, and IVIM simultaneous PWI/DWI ratio predicted infarct growth., Comment: Incomplete analysis

This paper presents a comprehensive analysis of storage proofs in the Ethereum ecosystem, examining their role in addressing historical and cross-chain state access challenges. We systematically review existing approaches to historical state verification, comparing Merkle Mountain Range (MMR) and Merkle-Patricia trie (MPT) architectures. An analysis involves their respective performance characteristics within zero-knowledge contexts, where performance challenges related to Keccak-256 are explored. The paper also examines the cross-chain verification, particularly focusing on the interactions between Ethereum and Layer 2 networks. Through careful analysis of storage proof patterns across different network configurations, we identify and formalize three architectures for cross-chain verification. By organizing this complex technical landscape, this analysis provides a structured framework for understanding storage proof implementations in the Ethereum ecosystem, offering insights into their practical applications and limitations., Comment: 22 pages, 15 figures

In Markov decision processes (MDPs), quantile risk measures such as Value-at-Risk are a standard metric for modeling RL agents' preferences for certain outcomes. This paper proposes a new Q-learning algorithm for quantile optimization in MDPs with strong convergence and performance guarantees. The algorithm leverages a new, simple dynamic program (DP) decomposition for quantile MDPs. Compared with prior work, our DP decomposition requires neither known transition probabilities nor solving complex saddle point equations and serves as a suitable foundation for other model-free RL algorithms. Our numerical results in tabular domains show that our Q-learning algorithm converges to its DP variant and outperforms earlier algorithms.

This paper presents a comprehensive framework aimed at enhancing education in modeling, optimal control, and nonlinear Model Predictive Control~(MPC) through a practical greenhouse climate control model. The framework includes a detailed mathematical model of lettuce growth and greenhouse, which are influenced by real-time external weather conditions obtained via an application programming interface~(API). Using this data, the MPC-based approach dynamically adjusts greenhouse conditions, optimizing plant growth and energy consumption and minimizing the social cost of CO\textsubscript{2}. The presented results demonstrate the effectiveness of this approach in balancing energy use with crop yield and reducing CO\textsubscript{2} emissions, contributing to economic efficiency and environmental sustainability. Besides optimizing lettuce production, the framework also provides a valuable resource for making control systems education more engaging and effective. The main aim is to provide students with a hands-on platform to understand the principles of modeling, the complexity of MPC and the trade-offs between profitability and sustainability in agricultural systems. This framework provides students with hands-on experience, helping them to understand the control theory better, connecting it to the practical implementation, and developing their problem-solving skills. The framework can be accessed at \url{ecompc4greenhouse.streamlit.app}., Comment: preprint under review in ECC25

This work presents steps toward the design of Majorana zero modes (MZM) in InAsP quantum dots (QD) embedded in an InP semiconducting nanowire in contact with a p-type superconductor described by the Kitaev Hamiltonian. The single particle spectrum is obtained from million atom atomistic calculations with QNANO and many-electron spectra using exact diagonalization (ED) and the hybrid Variational Quantum Eigensolver (VQE) method. A variational ansatz is constructed to capture the ground state of the system by utilizing a generalized form of the analytical solution for a particular set of parameters. By systematically deviating from the analytically solvable regime while maintaining the system in the topological phase (TP), the effectiveness of the variational function in reproducing the correct ground state and topological properties of the system is evaluated. This is done through a quantum algorithm for a many-body state containing MZM. The results are compared with exact solution in topological phase and demonstrate the capability of VQE, along with classical simulations, to accurately model the many-body spectra in topologically nontrivial state.

In this work, we present a first-principles density functional theory (DFT) computational investigation of the luminescence and absorption lineshapes associated with the neutral carbon-oxygen interstitial pair (CiOi) defect in silicon. We obtain the lineshapes of the defect in the dilute limit using a computational methodology that constructs dynamical matrices of supercells containing tens of thousands of atoms, utilizing systems directly accessible through DFT. Both perturbed bulk phonons and localized vibrations contribute to the phonon sideband. We achieve excellent agreement with experimental luminescence data. Our findings further reinforce the attribution of the well-known C-line in silicon to the neutral CiOi complex., Comment: 14 pages, 9 figures, 3 tables, 16 equations

Sherpa is a general-purpose Monte Carlo event generator for the simulation of particle collisions in high-energy collider experiments. We summarise new developments, essential features, and ongoing improvements within the Sherpa 3 release series. Physics improvements include higher-order electroweak corrections, simulations of photoproduction and hard diffraction at NLO QCD, heavy-flavour matching in NLO multijet merging, spin-polarised cross section calculations, and a new model of colour reconnections. In addition, the modelling of hadronisation, the underlying event and QED effects in both production and decay has been improved, and the overall event generation efficiency has been enhanced., Comment: We proudly present the new major release of the Sherpa MC generator

We develop a generic policy gradient method with the global optimality guarantee for robust Markov Decision Processes (MDPs). While policy gradient methods are widely used for solving dynamic decision problems due to their scalable and efficient nature, adapting these methods to account for model ambiguity has been challenging, often making it impractical to learn robust policies. This paper introduces a novel policy gradient method, Double-Loop Robust Policy Mirror Descent (DRPMD), for solving robust MDPs. DRPMD employs a general mirror descent update rule for the policy optimization with adaptive tolerance per iteration, guaranteeing convergence to a globally optimal policy. We provide a comprehensive analysis of DRPMD, including new convergence results under both direct and softmax parameterizations, and provide novel insights into the inner problem solution through Transition Mirror Ascent (TMA). Additionally, we propose innovative parametric transition kernels for both discrete and continuous state-action spaces, broadening the applicability of our approach. Empirical results validate the robustness and global convergence of DRPMD across various challenging robust MDP settings.

Large language models (LLMs) have demonstrated remarkable performance across various downstream tasks. However, the high computational and memory requirements of LLMs are a major bottleneck. To address this, parameter-efficient fine-tuning (PEFT) methods such as low-rank adaptation (LoRA) have been proposed to reduce computational costs while ensuring minimal loss in performance. Additionally, knowledge distillation (KD) has been a popular choice for obtaining compact student models from teacher models. In this work, we present KD-LoRA, a novel fine-tuning method that combines LoRA with KD. Our results demonstrate that KD-LoRA achieves performance comparable to full fine-tuning (FFT) and LoRA while significantly reducing resource requirements. Specifically, KD-LoRA retains 98% of LoRA's performance on the GLUE benchmark, while being 40% more compact. Additionally, KD-LoRA reduces GPU memory usage by 30% compared to LoRA, while decreasing inference time by 30% compared to both FFT and LoRA. We evaluate KD-LoRA across three encoder-only models: BERT, RoBERTa, and DeBERTaV3. Code is available at https://github.com/rambodazimi/KD-LoRA., Comment: Accepted at 4th NeurIPS Efficient Natural Language and Speech Processing Workshop (ENLSP-IV 2024)

The STS detector in the CBM experiment delivers data via multiple e-links connected to GBTX ASICs. In the process of data aggregation, that data must be received, combined into a smaller number of streams, and packed into so-called microslices containing data from specific periods. The aggregation must consider data randomization due to amplitude-dependent processing time in the FEE ASICs and different occupancy of individual e-links. During the development of the STS readout, the continued progress in the available technology affected the requirements for data aggregation, its architecture, and algorithms. The contribution presents considered solutions and discusses their properties., Comment: TWEPP 2024 Topical Workshop on Electronics for Particle Physics

We propose a $O(\log k \log n)$-competitive randomized algorithm for online node-weighted Steiner forest. This is essentially optimal and significantly improves over the previous bound of $O(\log^2 k \log n)$ by Hajiaghayi et al. [2017]. In fact, our result extends to the more general prize-collecting setting, improving over previous works by a poly-logarithmic factor. Our key technical contribution is a randomized online algorithm for set cover and non-metric facility location in a new adversarial model which we call semi-adaptive adversaries. As a by-product of our techniques, we obtain the first deterministic $O(\log |C| \log |F|)$-competitive algorithm for non-metric facility location., Comment: to appear in SODA 2025

The decomposition from the group theory-based methods of Sasao and Saraivanov is extended to design binary quantum cascades, using the quantum rotational gates by the X-axis (CNOT and RX), Y-axis (RY), and Z-axis (controlled-Z) of the Bloch sphere. A class of local transformations is also presented to simplify the final canonical cascade circuits. Our proposed methodology is well suited for quantum layouts, as each single-qubit gate has one target qubit and each double-qubit gate has one control qubit and one target qubit, thereby never creating a graph of triangular connectivity., Comment: 22 pages, 31 figures, 4 tables, Accepted to Journal of Applied Logics - IfCoLog Journal of Logics and their Applications

This paper extends the decomposition from the group theory based methods of Sasao and Saraivanov to design binary input multivalued output quantum cascades realized with optical NOT, SWAP, and Fredkin Gates. We present this method for 3, 5, and 7 valued outputs, but in general it can be used for odd prime valued outputs. The method can be extended to realize hybrid functions with different valued outputs. A class of local transformations is presented that can simplify the final cascade circuits. Using these simplifying transformations, we present an upper bound on the maximum number of gates in an arbitrary $n$-variable input and $k$-valued output function., Comment: 31 pages, 64 figures, Accepted to Quantum Information and Computation Journal

A generic Boolean-phase SWAP gate is introduced for quantum cost-effectiveness. This gate is termed the "p-SWAP", and p is a customizable phase difference between the swapped qubits, where 0 <= p <= ${\pm \pi}$ radians. The cost-effectiveness of p-SWAP gate comes from a lower quantum cost for its final transpiled quantum circuit into a real quantum computer. The p-SWAP gate only utilizes two Feynman (CNOT) gates, as compared to the standard SWAP gate constructed from three CNOT gates. The quantum circuit of p-SWAP gate is geometrically designed using our Bloch sphere approach. The generality of p-SWAP gate is proposed for Phase oracles requiring a desirable p for a combination set of swapped qubits, as well as for Boolean oracles requiring a cost-effective SWAP gate when p is ignored. In this paper, after transpilation (synthesization) into a real quantum computer, it was concluded that the transpiled quantum circuit of p-SWAP gate has a lower quantum cost than that of the standard SWAP gate., Comment: arXiv admin note: text overlap with arXiv:2410.13104

Orbital eccentricity in compact binary mergers carries crucial information about the binary's formation and environment. There are emerging signs that some of the mergers detected by the LIGO and Virgo gravitational wave detectors could indeed be eccentric. Nevertheless, the identification of eccentricity via gravitational waves remains challenging, to a large extent because of the limited availability of eccentric gravitational waveforms. While multiple suites of eccentric waveforms have recently been developed, they each cover only a part of the binary parameter space. Here we evaluate the sensitivity of LIGO to eccentric waveforms from the SXS and RIT numerical relativity catalogs and the TEOBResumS-Dali waveform model using data from LIGO-Virgo-Kagra's third observing run. The obtained sensitivities, as functions of eccentricity, mass and mass ratio, are important inputs to understanding detection prospects and observational population constrains. In addition, our results enable the comparison of the waveforms to establish their compatibility and applicability for searches and parameter estimation., Comment: 9 pages, 2 figures, 2 tables

Burst searches identify gravitational-wave (GW) signals in the detector data without use of a specific signal model, unlike the matched-filter searches that correlate data with simulated signal waveforms (templates). While matched filters are optimal for detection of known signals in the Gaussian noise, the burst searches can be more efficient in finding unusual events not covered by templates or those affected by non-Gaussian noise artifacts. Here, we report the detection of 3 gravitational wave signals that are uncovered by a burst search Coherent WaveBurst (cWB) optimized for the detection of binary black hole (BBH) mergers. They were found in the data from the LIGO/Virgo's third observing run (O3) with a combined significance of 3.6 $\sigma$. Each event appears to be a BBH merger not previously reported by the LIGO/Virgo's matched-filter searches. The most significant event has a reconstructed primary component in the upper mass gap ($m_1 = 70^{+36}_{-18}\,$M$_\odot$), and unusually low mass ratio ($m_2/m_1\sim0.3$), implying a dynamical or AGN origin. The 3 new events are consistent with the expected number of cWB-only detections in the O3 run ($4.8 \pm 2.1$), and belong to the stellar-mass binary population with the total masses in the $70-100$ M$_\odot$ range., Comment: 10 pages, 2 figures, 3 tables

This paper contains a study on stochastic Volterra integral equations with fuzzy sets-values and involving on a constant retardation. Moreover, the form of the equation is symmetric in the sense that fuzzy stochastic integrals are placed on both sides of the equation. We show that the considered initial value problem formulated in terms of symmetric fuzzy stochastic Volterra integral equation is well-posed. In particular, we show that there exists a unique solution and this solution depends continuously on the parameters of the equation. The results are achieved with the conditions of Lipschitz continuity of drift and diffusion coefficients, and continuity of kernels, Comment: 20 pages

In this research, we quantify an inflow of women into science in the past three decades. Structured Big Data allow us to estimate the contribution of women scientists to the growth of science by disciplines (N = STEMM 14 disciplines) and over time (1990-2023). A monolithic segment of STEMM science emerges from this research as divided between the disciplines in which the growth was powerfully driven by women - and the disciplines in which the role of women was marginal. There are four disciplines in which 50% of currently publishing scientists are women; and five disciplines in which more than 50% of currently young scientists are women. But there is also a cluster of four highly mathematized disciplines (MATH, COMP, PHYS, and ENG) in which the growth of science is only marginally driven by women. Digital traces left by scientists in their publications indexed in global datasets open two new dimensions in large-scale academic profession studies: time and gender. The growth of science in Europe was accompanied by growth in the number of women scientists, but with powerful cross-disciplinary and cross-generational differentiations. We examined the share of women scientists coming from ten different age cohorts for 32 European and four comparator countries (the USA, Canada, Australia, and Japan). Our study sample was N = 1,740,985 scientists (including 39.40% women scientists). Three critical methodological challenges of using structured Big Data of the bibliometric type were discussed: gender determination, academic age determination, and discipline determination., Comment: 27 pages (plus Supplementary Material 15 pages)

Self-supervised learning on tabular data seeks to apply advances from natural language and image domains to the diverse domain of tables. However, current techniques often struggle with integrating multi-domain data and require data cleaning or specific structural requirements, limiting the scalability of pre-training datasets. We introduce PORTAL (Pretraining One-Row-at-a-Time for All tabLes), a framework that handles various data modalities without the need for cleaning or preprocessing. This simple yet powerful approach can be effectively pre-trained on online-collected datasets and fine-tuned to match state-of-the-art methods on complex classification and regression tasks. This work offers a practical advancement in self-supervised learning for large-scale tabular data., Comment: Accepted at Table Representation Learning Workshop at NeurIPS 2024

The skull segmentation from CT scans can be seen as an already solved problem. However, in MR this task has a significantly greater complexity due to the presence of soft tissues rather than bones. Capturing the bone structures from MR images of the head, where the main visualization objective is the brain, is very demanding. The attempts that make use of skull stripping seem to not be well suited for this task and fail to work in many cases. On the other hand, supervised approaches require costly and time-consuming skull annotations. To overcome the difficulties we propose a fully unsupervised approach, where we do not perform the segmentation directly on MR images, but we rather perform a synthetic CT data generation via MR-to-CT translation and perform the segmentation there. We address many issues associated with unsupervised skull segmentation including the unpaired nature of MR and CT datasets (contrastive learning), low resolution and poor quality (super-resolution), and generalization capabilities. The research has a significant value for downstream tasks requiring skull segmentation from MR volumes such as craniectomy or surgery planning and can be seen as an important step towards the utilization of synthetic data in medical imaging., Comment: 16 pages, 5 figures, ACCV 2024 - GAISynMeD Workshop

A cost-effective n-bit Toffoli gate is proposed to be realized (or transpiled) based on the layouts (linear, T-like, and I-like) and the number of n physical qubits for IBM quantum computers. This proposed gate is termed the "layout-aware n-bit Toffoli gate". The layout-aware n-bit Toffoli gate is designed using the visual approach of the Bloch sphere, from the visual representations of the rotational quantum operations for IBM native gates. In this paper, we also proposed a new formula for the quantum cost, which calculates the total number of native gates, the crossing connections, and the depth of the final transpiled quantum circuit. This formula is termed the "transpilation quantum cost". After transpilation, our proposed layout-aware n-bit Toffoli gate always has a much lower transpilation quantum cost than that of the conventional n-bit Toffoli gate, where 3 <= n <= 7 qubits, for different IBM quantum computers., Comment: 22 figures and 5 tables. arXiv admin note: text overlap with arXiv:2311.06760

Monolayer transition metal dichalcogenides have emerged as prominent candidates to explore the complex interplay between the spin and the valleys degrees of freedom. The strong spin-orbit interaction and broken inversion symmetry within these materials lead to the spin-valley locking effect, in which carriers occupying the K and K' valleys of the reciprocal space must have opposite spin depending on which valley they reside. This effect is particularly strong for holes due to a larger spin-orbit gap in the valence band. By reducing the dimensionality of a monolayer of tungsten diselenide to 1D via electrostatic confinement, we demonstrate that spin-valley locking in combination with strong hole-hole interactions lead to a ferromagnetic state in which hole transport through the 1D system is spin-valley polarized, even without an applied magnetic field, and that the persistence of this spin-valley polarized configuration can be tuned by a global back-gate. This observation opens the possibility of implementing a robust and stable valley polarized system, essential for valleytronic applications.

The TraX Engine is an advanced data processing tool developed by ADVACAM in collaboration with the European Space Agency (ESA), specifically designed for analyzing data from Timepix detectors equipped with various sensor materials (Si, CdTe, GaAs, SiC). TraX Engine can process large datasets across various scientific and medical applications, including space radiation monitoring, particle therapy, and imaging. In space applications, the TraX Engine has been used to process data from satellites like OneWeb JoeySat deployed in LEO orbit, where it continuously monitors space radiation environments measuring flux, dose, and dose rate in real-time. In medical applications, particularly in particle therapy, the TraX Engine is used to process data to characterize radiation fields in terms of particle flux, Linear Energy Transfer, and spatial distribution of the radiation dose. The TraX Engine can identify and classify scattered particles, such as secondary protons and electrons, and estimate their contribution to out-of-field doses. In imaging applications, the TraX Engine is integrated into Compton cameras, where it supports photon source localization through directional reconstruction of photons. The system ability to identify gamma radiation source with high precision makes it suitable for medical imaging tasks, such as tracking I-131 used in thyroid cancer treatment or localizing radiation sources. This paper presents the architecture and capabilities of the newly developed software TraX Engine, alongside results from various applications, demonstrating its role in particle tracking, radiation monitoring, imaging, and others. With its modular architecture, the TraX Engine offers multiple interfaces, including a command-line tool, an API, a web portal, and a graphical user interface, ensuring usability across different fields and user expertise levels., Comment: 14 pages, 7 figures

The financial industry has undergone a significant transition from the London Interbank Offered Rate (LIBOR) to Risk Free Rates (RFR) such as, e.g., the Secured Overnight Financing Rate (SOFR) in the U.S. and the AUD Overnight Index Average (AONIA) in Australia, as the primary benchmark rate for borrowing costs. The paper examines the pricing and hedging method for SOFR-related financial products in a cross-currency context with the special emphasis on the Compound SOFR vs Average AONIA cross-currency basis swaps. While the SOFR and AONIA serve as a particular case of a cross-currency basis swap (CCBS), the approach developed is able to handle backward-looking term rates for any two currencies. We give explicit pricing and hedging results for collateralized cross-currency basis swaps using interest rate and currency futures contracts as hedging tools within an arbitrage-free multi-curve setting., Comment: 46 pages, 6 gifures

We consider the spin-orbit coupled Bose gas with repulsive mean-field interparticle interactions. We analyze the phase diagram of the system varying the temperature $T>0$, the chemical potentials, as well as interparticle and spin-orbit interaction couplings. Our results indicate that, for Rashba- and Weyl-type spin-orbit couplings, condensates featuring ordering wavevector $\vec{Q}\neq \vec{0}$ are fragile with respect to thermal fluctuations and, at $T>0$, the only stable thermodynamic phases involving the Bose-Einstein condensate (BEC) are those of uniform type with $\vec{Q}=\vec{0}$. On the other hand, presence of the spin-orbit coupling stabilizes the $\vec{Q}=\vec{0}$ BEC state at any dimensionality $d>1$ and modifies either the order or the universality class of the corresponding phase transition. We emphasize the singular nature of the limit of vanishing spin-orbit interaction coupling $v$, sizable shifts of the phase boundaries upon varying $v$, as well as the role of the relative magnitudes of the interparticle interaction couplings for the character of the condensation transition.

Radiation from the "plunging region" (the space between the innermost stable circular orbit (ISCO) and the black-hole (BH) surface) of an accretion flow onto a BH is supposed to add a small but significant contribution to the X-ray spectra of X-ray binary systems. The plunging region and its electromagnetic emission has been recently described by numerical and analytic calculations which lead to the conclusion that in the plunging region radiation is generated by the energy release through the action of a "leftover" stress in the vicinity of the ISCO and that the amount of this leftover can be chosen as a free parameter of the accretion-flow description. The present article aims to demonstrate that this is not true because the stress in the whole accretion flow onto a black hole is determined by its global transonic character enforced by the space-time structure of the accreting black hole. We use the slim-disc approximation (SDA) to illustrate this point. In our article we compare models obtained with the SDA with results of numerical simulations and of analytical models based on the assumption that accreting matter flows along geodesics. We show that the latter cannot describe adequately the physics of astrophysical accretion onto a BH because 1. particles on geodesics cannot emit electromagnetic radiation, 2. they ignore the global transonic character of the accretion flow imposed by the presence of a stationary horizon in the BH spacetime; a presence that fixes a unique value of the angular momentum at the BH surface for a solution to exist. Therefore the fact that geodesics-based models reproduce the trans-ISCO behaviour of GRRMHD simulations proves the physical reality of neither. We show also that the claimed detection of plunging-region emission in the spectrum of an X-ray binary is unsubstantiated., Comment: Submitted to Astronomy and Astrophysics, comments welcome