Search

Recent advances in machine learning have opened new avenues for optimizing detector designs in high-energy physics, where the complex interplay of geometry, materials, and physics processes has traditionally posed a significant challenge. In this work, we introduce the $\textit{end-to-end}$ optimization framework AIDO that leverages a diffusion model as a surrogate for the full simulation and reconstruction chain, enabling gradient-based design exploration in both continuous and discrete parameter spaces. Although this framework is applicable to a broad range of detectors, we illustrate its power using the specific example of a sampling calorimeter, focusing on charged pions and photons as representative incident particles. Our results demonstrate that the diffusion model effectively captures critical performance metrics for calorimeter design, guiding the automatic search for layer arrangement and material composition that aligns with known calorimeter principles. The success of this proof-of-concept study provides a foundation for future applications of end-to-end optimization to more complex detector systems, offering a promising path toward systematically exploring the vast design space in next-generation experiments., Comment: 14 pages, 9 figures, submitted to MDPI particles

Transition metal oxides are a platform for exploring strain-engineered intriguing physical properties and developing spintronic or flexible electronic functionalities owing to strong coupling of spin, charge and lattice degrees of freedom. In this study, we exemplify the strain-engineered magnetism of La$_{2/3}$Sr$_{1/3}$MnO$_3$ in freestanding and rippled membrane forms without and with process-induced strain, respectively, prepared by epitaxial lift-off technique. We find that the deposition of Pt/Ti stressor suppresses the crack formation in the lift-off process and induces a ripple structure in the La$_{2/3}$Sr$_{1/3}$MnO$_3$ membrane. Laser micrograph and Raman spectroscopy show a ripple period of about 30 um and a height of a few um, where alternating convex and concave structures are subjected to tensile strain of 0.6% and compressive strain of 0.5%, respectively. While the freestanding La$_{2/3}$Sr$_{1/3}$MnO$_3$ membrane exhibits room-temperature ferromagnetism, the macroscopic magnetic transition temperature (TC) of the rippled membrane is reduced by as large as 27%. Temperature-variable Kerr microscopy observation in the rippled membrane reveals that the spatial variation of TC to be approximately 4% of the macroscopic TC, which coincides with the local strains at convex and concave structures. The large reduction of macroscopic TC in the rippled membrane may be ascribed to the lattice disorders due to strain gradient. Our demonstration of tuning ferromagnetism by the ripple structure validates the high potential of the process-induced strain in epitaxial lift-off technique and paves the way for strain-mediated emerging physical properties in various transition metal oxides.

We perform fully general-relativistic hydrodynamics simulations of binary neutron star mergers over $100\,\rm ms$ post-merger to investigate the dynamics of remnant massive neutron stars (NSs). Our focus is mainly on the analysis of convective stability and mode characteristics of the massive NSs. We derive stability criteria for hot, differentially rotating relativistic stars that account for both buoyant and rotational restoring forces, and apply them for the first time to the post-merger massive NSs. Our results show no evidence of large-scale convective instability, as both angle-averaged specific entropy and specific angular momentum increase outward within the massive NSs. Rotational effects significantly enhance stability for local regions that would be otherwise unstable by the Schwarzschild criterion. Additionally, our mode analysis of matter fields and gravitational waves reveals no excitation of inertial modes after the damping of quadrupolar $f$-modes in the massive NSs, contrasting with previous studies. As in many previous works, we observe the excitation of an $m=1$ one-armed mode. However, we also find that the growth of the $m=1$ mode amplitude after the merger may correlate strongly with the violation of linear momentum conservation, indicating that we cannot reject the possibility that the excitation of the one-armed mode has a numerical origin., Comment: 22 pages, 18 figures; submitted to Phys. Rev. D

In this paper, we investigate the asymptotic behavior of small solutions to the initial value problem for a system of cubic nonlinear Schrodinger equations (NLS) in one spatial dimension. We identify a new class of NLS systems for which the global boundedness and asymptotics of small solutions can be established, even in the absence of any effective conserved quantity. The key to this analysis lies in utilizing conserved quantities for the reduced ordinary differential equation (ODE) systems derived from the original NLS systems. In a previous study, the first author investigated conserved quantities expressed as quartic polynomials. In contrast, the conserved quantities considered in the present paper are of a different type and are not necessarily polynomial.

For large-scale network systems, network centrality based on control theory plays a crucial role in understanding their properties and controlling them efficiently. The controllability score is such a centrality index and can give a physically meaningful measure. Nevertheless, the existing work is limited to linear time-invariant (LTI) systems and the controllability score cannot be applied to linear time-varying (LTV) systems, which include essential models such as temporal networks for real application. This paper extends it to apply to LTV systems. Since it is defined as an optimal solution to some optimization problem, it is not necessarily uniquely determined. Its uniqueness must be guaranteed for reproducibility and interpretability. This paper also shows its uniqueness in most practical cases, which guarantees its use as a network centrality. In addition, we propose a data-driven method to compute it for its practical use. Finally, in numerical experiments, we compare controllability scores between LTI and LTV systems and assess the performance of the proposed data-driven method.

We study the problem of online clustering within the multi-armed bandit framework under the fixed confidence setting. In this multi-armed bandit problem, we have $M$ arms, each providing i.i.d. samples that follow a multivariate Gaussian distribution with an {\em unknown} mean and a known unit covariance. The arms are grouped into $K$ clusters based on the distance between their means using the Single Linkage (SLINK) clustering algorithm on the means of the arms. Since the true means are unknown, the objective is to obtain the above clustering of the arms with the minimum number of samples drawn from the arms, subject to an upper bound on the error probability. We introduce a novel algorithm, Average Tracking Bandit Online Clustering (ATBOC), and prove that this algorithm is order optimal, meaning that the upper bound on its expected sample complexity for given error probability $\delta$ is within a factor of 2 of an instance-dependent lower bound as $\delta \rightarrow 0$. Furthermore, we propose a computationally more efficient algorithm, Lower and Upper Confidence Bound-based Bandit Online Clustering (LUCBBOC), inspired by the LUCB algorithm for best arm identification. Simulation results demonstrate that the performance of LUCBBOC is comparable to that of ATBOC. We numerically assess the effectiveness of the proposed algorithms through numerical experiments on both synthetic datasets and the real-world MovieLens dataset. To the best of our knowledge, this is the first work on bandit online clustering that allows arms with different means in a cluster and $K$ greater than 2.

Carbon Monoxide (CO) is a dominant pollutant in urban areas due to the energy generation from fossil fuels for industry, automobile, and domestic requirements. Forecasting the evolution of CO in real-time can enable the deployment of effective early warning systems and intervention strategies. However, the computational cost associated with the physics and chemistry-based simulation makes it prohibitive to implement such a model at the city and country scale. To address this challenge, here, we present a machine learning model based on neural operator, namely, Complex Neural Operator for Air Quality (CoNOAir), that can effectively forecast CO concentrations. We demonstrate this by developing a country-level model for short-term (hourly) and long-term (72-hour) forecasts of CO concentrations. Our model outperforms state-of-the-art models such as Fourier neural operators (FNO) and provides reliable predictions for both short and long-term forecasts. We further analyse the capability of the model to capture extreme events and generate forecasts in urban cities in India. Interestingly, we observe that the model predicts the next hour CO concentrations with R2 values greater than 0.95 for all the cities considered. The deployment of such a model can greatly assist the governing bodies to provide early warning, plan intervention strategies, and develop effective strategies by considering several what-if scenarios. Altogether, the present approach could provide a fillip to real-time predictions of CO pollution in urban cities., Comment: 28 pages, 14 figures, under submission process

We generally classify the equivalence classes of the $T^2/Z_m$ $(m=2,3,4,6)$ orbifold boundary conditions (BCs) for the $SO(N)$ gauge group. Higher-dimensional gauge theories are defined by gauge groups, matter field contents, and the BCs. The numerous patterns of the BCs are classified into the finite equivalence classes, each of which consists of the physically equivalent BCs. In this paper, we reconstruct the canonical forms of the BCs for the $SO(N)$ gauge group through the ``re-orthogonalization method." All the possible equivalent relations between the canonical forms are examined by using the trace conservation laws. The number of the equivalence classes in each orbifold model is obtained., Comment: 25 pages, 2 figures

The Mice Autism Detection via Ultrasound Vocalization (MADUV) Challenge introduces the first INTERSPEECH challenge focused on detecting autism spectrum disorder (ASD) in mice through their vocalizations. Participants are tasked with developing models to automatically classify mice as either wild-type or ASD models based on recordings with a high sampling rate. Our baseline system employs a simple CNN-based classification using three different spectrogram features. Results demonstrate the feasibility of automated ASD detection, with the considered audible-range features achieving the best performance (UAR of 0.600 for segment-level and 0.625 for subject-level classification). This challenge bridges speech technology and biomedical research, offering opportunities to advance our understanding of ASD models through machine learning approaches. The findings suggest promising directions for vocalization analysis and highlight the potential value of audible and ultrasound vocalizations in ASD detection., Comment: 5 pages, 1 figure and 2 tables. For MADUV Challenge 2025

We directly evaluate the probability amplitudes in Jackiw-Teitelboim (JT) gravity using the Lorentzian path integral formulation. By imposing boundary conditions on the scale factor and the dilaton field, the Lorentzian path integral uniquely yields the probability amplitude without contradiction. Under Dirichlet boundary conditions, we demonstrate that the amplitude derived from the Lorentzian path integral is expressed in terms of the modified Bessel function of the second kind. Furthermore, we provide the determinant for various boundary conditions and perform a detailed analysis of the Lefschetz thimble structure and saddle points. In contrast to four-dimensional gravity, we show that the Hartle-Hawking no-boundary proposal is approximately valid in JT quantum cosmology. Furthermore, addressing quantum perturbation issues, we show that the quantum genesis of the two-dimensional universe occurs and exhibits perturbative regularity when the dilaton field is non-zero and large as an initial condition., Comment: 18 pages, 4 figures

In this work, we combine conventional linear response time-domain THz spectroscopy with non-linear THz-pump THz-probe techniques to study metallic strained thin films of $\mathrm{Ca}_2\mathrm{RuO}_4$, which undergo a transition into a ferromagnetic state at 10 K. Such measurements allowing us to independently measure momentum and energy relaxation rates. We find that while the momentum relaxation rate decreases significantly at the ferromagnetic transition, the energy relaxation rate remains unaffected by the emergence of magnetic order. This shows that the dominant changes to scattering across the transition correspond to scatterings that relax momentum without relaxing energy. It is consistent with a scenario where energy is not carried off by coupling to collective magnetic degrees of freedom. Instead, the principal channel for energy relaxation remains the conventional one e.g. coupling to acoustic phonons. This observation validates the approximation used in the conventional understanding of resistive anomalies of ferromagnets across the Curie temperature, which due to critical slowing down, spin fluctuations can be treated as effectively static and scattering off of them elastic. This scenario can likely be extended to resistive anomalies at other phase transitions to charge- and spin-density wave states in kagome metals or pnictide system, Comment: 6 pages, 4 figures

A real number is called simply normal to base $b$ if every digit $0,1,\ldots ,b-1$ should appear in its $b$-adic expansion with the same frequency $1/b$. A real number is called normal to base $b$ if it is simply normal to every base $b, b^2, \ldots$. In this article, we discover a relation between the normality of algebraic numbers and a mean of the Riemann zeta function on vertical arithmetic progressions. Consequently, we reveal that a positive algebraic irrational number $\alpha$ is normal to base $b$ if and only if we have \[ \lim_{N\to \infty}\frac{1}{\log N} \sum_{1\leq |n|\leq N} \zeta\left(-k+\frac{2\pi i n}{\log b} \right) \frac{e^{2\pi i n \log \alpha /\log b}}{n^{k+1}} =0 \] for every integer $k\geq 0$., Comment: 29 pages

Grover's quantum algorithm can find a marked item from an unstructured database faster than any classical algorithm, and hence it has been used for several applications such as cryptanalysis and optimization. When there exist multiple marked items, Grover's algorithm has the property of finding one of them uniformly at random. To further broaden the application range, it was generalized so that it finds marked items with probabilities according to their priority by encoding the priority into amplitudes applied by Grover's oracle operator. In this paper, to achieve a similar generalization, we examine a different encoding that incorporates the priority into phases applied by the oracle operator. We compare the previous and our oracle operators and observe that which one performs better depends on priority parameters. Since the priority parameters can be considered as the magnitude of the correlated phase error on Grover's oracle operator, the analysis of our oracle operator also reveals the robustness of the original Grover's algorithm against correlated noises. We further numerically show that the coherence between multiple marked items increases the probability of finding the most prioritized one in Grover's algorithm with our oracle operator., Comment: 20 pages, 9 figures

In recent years, labor shortages due to the declining birthrate and aging population have become significant challenges at construction sites in developed countries, including Japan. To address these challenges, we are developing an open platform called ROS2-TMS for Construction, a Cyber-Physical System (CPS) for construction sites, to achieve both efficiency and safety in earthwork operations. In ROS2-TMS for Construction, the system comprehensively collects and stores environmental information from sensors placed throughout the construction site. Based on these data, a real-time virtual construction site is created in cyberspace. Then, based on the state of construction machinery and environmental conditions in cyberspace, the optimal next actions for actual construction machinery are determined, and the construction machinery is operated accordingly. In this project, we decided to use the Open Platform for Earthwork with Robotics and Autonomy (OPERA), developed by the Public Works Research Institute (PWRI) in Japan, to control construction machinery from ROS2-TMS for Construction with an originally extended behavior tree. In this study, we present an overview of OPERA, focusing on the newly developed navigation package for operating the crawler dump, as well as the overall structure of ROS2-TMS for Construction as a Cyber-Physical System (CPS). Additionally, we conducted experiments using a crawler dump and a backhoe to verify the aforementioned functionalities.

While recent text-to-image models can generate photorealistic images from text prompts that reflect detailed instructions, they still face significant challenges in accurately rendering words in the image. In this paper, we propose to retouch erroneous text renderings in the post-processing pipeline. Our approach, called Type-R, identifies typographical errors in the generated image, erases the erroneous text, regenerates text boxes for missing words, and finally corrects typos in the rendered words. Through extensive experiments, we show that Type-R, in combination with the latest text-to-image models such as Stable Diffusion or Flux, achieves the highest text rendering accuracy while maintaining image quality and also outperforms text-focused generation baselines in terms of balancing text accuracy and image quality.

We investigate a pseudo-Nambu-Goldstone (pNG) dark matter (DM) model based on a gauged $SU(2)_x$ and a global $SU(2)_g$ symmetries. These symmetries are spontaneously broken to a global $U(1)_D$ symmetry by a vacuum expectation value of an $SU(2)_x \times SU(2)_g$ bi-fundamental scalar field. The global $SU(2)_g$ symmetry is also softly broken to a global $U(1)_D$ symmetry. Under the setup, a complex pNG boson arises. It is stabilized by $U(1)_D$ and is a DM candidate. Its scattering cross section off a nucleon is highly suppressed by small momentum transfer and thus evades the stringent constraints from DM direct detection experiments. Assuming all the couplings in the dark sector are real, a discrete symmetry arises. Consequently, in addition to the pNG DM, the lighter one of an $SU(2)_x$ gauge boson $V^0$ and a CP-odd scalar boson $a_0$ from the bi-fundamental scalar field can also serve as a DM candidate. Therefore, the model provides two-component DM scenarios. We find that the relic abundance of the DM candidates explains the measured value of the DM energy density. We also find that the pNG DM is the dominant DM component in large regions of the parameter space. In contrast to the pNG DM, both $V^0$ and $a_0$ scatter off a nucleon, and their scattering cross sections are not suppressed. However, their scattering event rates are suppressed by their number densities. Thus, the scattering cross section is effectively reduced. We show that the effective WIMP-nucleon scattering cross sections in the two-component scenarios are smaller than the current upper bounds and above the neutrino fog., Comment: 36 pages, 13 figures

Silicon Photonics-based AI Accelerators (SPAAs) have been considered as promising AI accelerators achieving high energy efficiency and low latency. While many researchers focus on improving SPAAs' energy efficiency and latency, their physical security has not been sufficiently studied. This paper first proposes a threat of thermal fault injection attacks on SPAAs based on Vector-Matrix Multipliers (VMMs) utilizing Mach-Zhender Interferometers. This paper then proposes SecONN, an optical neural network framework that is capable of not only inferences but also concurrent detection of the attacks. In addition, this paper introduces a concept of Wavelength Division Perturbation (WDP) where wavelength dependent VMM results are utilized to increase detection accuracy. Simulation results show that the proposed method achieves 88.7% attack-caused average misprediction recall.

We examine the vacuum state and its corresponding renormalized stress-energy tensor (RSET) in static horizonless regular spacetime in both two- and four-dimensions. Using the local field formulation of the anomaly-induced effective action, we show that the regularities of the spacetime and the RSET dictate the appropriate vacuum state. Furthermore, through a case study under the horizonless Bardeen-type spacetime, we demonstrate that the preferred vacuum state is not the Boulware vacuum, but a non-trivial one with a different RSET profile., Comment: 19 pages, 15 figures

We have investigated terahertz (THz) nonlinear responses in a multi-gap superconductor, MgB$_2$, using THz two-dimensional coherent spectroscopy (THz 2DCS). With broad-band THz drives, we identified a well-defined nonlinear response near the lower superconducting gap energy $2\Delta_{\pi}$ only at the lowest temperatures. Using narrow-band THz driving pulses, we observed first (FH) and third harmonic responses, and the FH intensity shows a monotonic increase with decreasing temperature when properly normalized by the driving field strength. This is distinct from the single-gap superconductor NbN, where the FH signal exhibited a resonant enhancement at temperatures near the superconducting transition temperature $T_{\text{c}}$ when the superconducting gap energy was resonant with the driving photon energy and which had been interpreted to originate from the superconducting amplitude mode. Our results in MgB$_2$ are consistent with a well-defined amplitude mode only at the lowest temperatures and indicate strong damping as temperature increases. This likely indicates the importance of interband coupling in MgB$_2$ and its influence on the nature of the amplitude mode and its damping.

The domain adaptation of language models, including large language models (LLMs), has become increasingly important as the use of such models continues to expand. This study demonstrates the effectiveness of Composition to Augment Language Models (CALM) in adapting to the financial domain. CALM is a model to extend the capabilities of existing models by introducing cross-attention between two LLMs with different functions. In our experiments, we developed a CALM to enhance the financial performance of an LLM with strong response capabilities by leveraging a financial-specialized LLM. Notably, the CALM was trained using a financial dataset different from the one used to train the financial-specialized LLM, confirming CALM's ability to adapt to various datasets. The models were evaluated through quantitative Japanese financial benchmarks and qualitative response comparisons, demonstrating that CALM enables superior responses with higher scores than the original models and baselines. Additionally, comparative experiments on connection points revealed that connecting the middle layers of the models is most effective in facilitating adaptation to the financial domain. These findings confirm that CALM is a practical approach for adapting LLMs to the financial domain.

This paper proposes a foundation model called "CLaSP" that can search time series signals using natural language that describes the characteristics of the signals as queries. Previous efforts to represent time series signal data in natural language have had challenges in designing a conventional class of time series signal characteristics, formulating their quantification, and creating a dictionary of synonyms. To overcome these limitations, the proposed method introduces a neural network based on contrastive learning. This network is first trained using the datasets TRUCE and SUSHI, which consist of time series signals and their corresponding natural language descriptions. Previous studies have proposed vocabularies that data analysts use to describe signal characteristics, and SUSHI was designed to cover these terms. We believe that a neural network trained on these datasets will enable data analysts to search using natural language vocabulary. Furthermore, our method does not require a dictionary of predefined synonyms, and it leverages common sense knowledge embedded in a large-scale language model (LLM). Experimental results demonstrate that CLaSP enables natural language search of time series signal data and can accurately learn the points at which signal data changes.

Machine Learning and data mining techniques (i.e. supervised and unsupervised techniques) are used across domains to detect user safety violations. Examples include classifiers used to detect whether an email is spam or a web-page is requesting bank login information. However, existing ML/DM classifiers are limited in their ability to understand natural languages w.r.t the context and nuances. The aforementioned challenges are overcome with the arrival of Gen-AI techniques, along with their inherent ability w.r.t translation between languages, fine-tuning between various tasks and domains. In this manuscript, we provide a comprehensive overview of the various work done while using Gen-AI techniques w.r.t user safety. In particular, we first provide the various domains (e.g. phishing, malware, content moderation, counterfeit, physical safety) across which Gen-AI techniques have been applied. Next, we provide how Gen-AI techniques can be used in conjunction with various data modalities i.e. text, images, videos, audio, executable binaries to detect violations of user-safety. Further, also provide an overview of how Gen-AI techniques can be used in an adversarial setting. We believe that this work represents the first summarization of Gen-AI techniques for user-safety.

We present HourVideo, a benchmark dataset for hour-long video-language understanding. Our dataset consists of a novel task suite comprising summarization, perception (recall, tracking), visual reasoning (spatial, temporal, predictive, causal, counterfactual), and navigation (room-to-room, object retrieval) tasks. HourVideo includes 500 manually curated egocentric videos from the Ego4D dataset, spanning durations of 20 to 120 minutes, and features 12,976 high-quality, five-way multiple-choice questions. Benchmarking results reveal that multimodal models, including GPT-4 and LLaVA-NeXT, achieve marginal improvements over random chance. In stark contrast, human experts significantly outperform the state-of-the-art long-context multimodal model, Gemini Pro 1.5 (85.0% vs. 37.3%), highlighting a substantial gap in multimodal capabilities. Our benchmark, evaluation toolkit, prompts, and documentation are available at https://hourvideo.stanford.edu, Comment: NeurIPS 2024 Datasets and Benchmarks Track; 28 pages

This paper proposes a framework of explaining anomalous machine sounds in the context of anomalous sound detection~(ASD). While ASD has been extensively explored, identifying how anomalous sounds differ from normal sounds is also beneficial for machine condition monitoring. However, existing sound difference captioning methods require anomalous sounds for training, which is impractical in typical machine condition monitoring settings where such sounds are unavailable. To solve this issue, we propose a new strategy for explaining anomalous differences that does not require anomalous sounds for training. Specifically, we introduce a framework that explains differences in predefined timbre attributes instead of using free-form text captions. Objective metrics of timbre attributes can be computed using timbral models developed through psycho-acoustical research, enabling the estimation of how and what timbre attributes have changed from normal sounds without training machine learning models. Additionally, to accurately determine timbre differences regardless of variations in normal training data, we developed a method that jointly conducts anomalous sound detection and timbre difference estimation based on a k-nearest neighbors method in an audio embedding space. Evaluation using the MIMII DG dataset demonstrated the effectiveness of the proposed method.

Kitaev quantum spin liquids (KSLs) in layered honeycomb magnets are known to host Majorana quasiparticles, whose excitations depend strongly on the direction of the applied magnetic field. In the high-field phase of $\alpha$-RuCl$_3$, specific heat measurements have revealed characteristic field-angle dependence of low-energy excitations consistent with the Kitaev model, providing bulk evidence for the KSL state. Here we present low-temperature measurements of specific heat $C(T)$ for another KSL candidate $\textrm{Na}_2\textrm{Co}_2\textrm{Te}\textrm{O}_6$ (NCTO) under field rotation within the honeycomb plane. Above the critical field of antiferromagnetic order, the field-angle dependence of $C/T$ exhibits minima along the bond directions, contrasting with the maxima observed in the KSL state of $\alpha$-RuCl$_3$. Our analysis indicates nodeless, fully-gapped excitations, which are inconsistent with the angle-dependent Majorana excitations with gapless nodes predicted by the Kitaev model. These findings suggest that low-energy excitations in NCTO are governed by gapped magnon excitations rather than Majorana quasiparticles, providing thermodynamic evidence against a KSL state., Comment: 6 pages, 4 figures

Forecasting rail congestion is crucial for efficient mobility in transport systems. We present rail congestion forecasting using reports from passengers collected through a transit application. Although reports from passengers have received attention from researchers, ensuring a sufficient volume of reports is challenging due to passenger's reluctance. The limited number of reports results in the sparsity of the congestion label, which can be an issue in building a stable prediction model. To address this issue, we propose a semi-supervised method for congestion forecasting for trains, or SURCONFORT. Our key idea is twofold: firstly, we adopt semi-supervised learning to leverage sparsely labeled data and many unlabeled data. Secondly, in order to complement the unlabeled data from nearby stations, we design a railway network-oriented graph and apply the graph to semi-supervised graph regularization. Empirical experiments with actual reporting data show that SURCONFORT improved the forecasting performance by 14.9% over state-of-the-art methods under the label sparsity., Comment: Accepted in ACM SIGSPATIAL 2024

Gen-AI techniques are able to improve understanding of context and nuances in language modeling, translation between languages, handle large volumes of data, provide fast, low-latency responses and can be fine-tuned for various tasks and domains. In this manuscript, we present a comprehensive overview of the applications of Gen-AI techniques in the finance domain. In particular, we present the opportunities and challenges associated with the usage of Gen-AI techniques. We also illustrate the various methodologies which can be used to train Gen-AI techniques and present the various application areas of Gen-AI technologies in the finance ecosystem. To the best of our knowledge, this work represents the most comprehensive summarization of Gen-AI techniques within the financial domain. The analysis is designed for a deep overview of areas marked for substantial advancement while simultaneously pin-point those warranting future prioritization. We also hope that this work would serve as a conduit between finance and other domains, thus fostering the cross-pollination of innovative concepts and practices.

We study the Dirichlet energy of some smooth maps appearing in a collapsing family of hyper-K\"ahler metrics on the $K3$ surface constructed by Foscolo. We introduce an invariant for homotopy classes of smooth maps from the $K3$ surface with a hyper-K\"ahler metric to a flat Riemannian orbifold of dimension $3$, then show that it gives a lower bound of the energy. Moreover, we show that the ratio of the energy to the invariant converges to $1$ for Foscolo's collapsing families.

Causal networks are widely used in many fields to model the complex relationships between variables. A recent approach has sought to construct causal networks by leveraging the wisdom of crowds through the collective participation of humans. While this can yield detailed causal networks that model the underlying phenomena quite well, it requires a large number of individuals with domain understanding. We adopt a different approach: leveraging the causal knowledge that large language models, such as OpenAI's GPT-4, have learned by ingesting massive amounts of literature. Within a dedicated visual analytics interface, called CausalChat, users explore single variables or variable pairs recursively to identify causal relations, latent variables, confounders, and mediators, constructing detailed causal networks through conversation. Each probing interaction is translated into a tailored GPT-4 prompt and the response is conveyed through visual representations which are linked to the generated text for explanations. We demonstrate the functionality of CausalChat across diverse data contexts and conduct user studies involving both domain experts and laypersons.

We performed the longest numerical-relativity neutrino-radiation magnetohydrodynamics simulation for a binary neutron star merger that extends to $\approx1.5\mathrm{\,s}$ after the merger. We consider the binary model that undergoes the prompt collapse to a black hole after the merger with asymmetric mass 1.25$\,M_{\odot}$ and 1.65$\,M_{\odot}$ and SFHo equation of state. We find the Poynting flux-driven collimated outflow as well as the gravitational wave emission, neutrino emission, dynamical mass ejection, and post-merger mass ejection facilitated by magnetorotational instability-driven turbulent viscosity in a single self-consistent binary neutron star merger simulation. A magnetosphere dominated by the aligned global magnetic field penetrating the black hole develops along the black-hole spin axis after the turbulence in the remnant disk is enhanced. A jet with the Poynting flux with isotropic-equivalent luminosity of $\sim10^{49}\mathrm{\,erg/s}$ is launched, and the duration of the high luminosity is expected to be $O(1)\mathrm{\,s}$., Comment: 14 pages, 9 figures

Hospitals in India still rely on handwritten medical records despite the availability of Electronic Medical Records (EMR), complicating statistical analysis and record retrieval. Handwritten records pose a unique challenge, requiring specialized data for training models to recognize medications and their recommendation patterns. While traditional handwriting recognition approaches employ 2-D LSTMs, recent studies have explored using Multimodal Large Language Models (MLLMs) for OCR tasks. Building on this approach, we focus on extracting medication names and dosages from simulated medical records. Our methodology MIRAGE (Multimodal Identification and Recognition of Annotations in indian GEneral prescriptions) involves fine-tuning the QWEN VL, LLaVA 1.6 and Idefics2 models on 743,118 high resolution simulated medical record images-fully annotated from 1,133 doctors across India. Our approach achieves 82% accuracy in extracting medication names and dosages., Comment: 5 pages, 9 figures, 3 tables, submitted to ISBI 2025

This paper proposes a zero-shot speech emotion recognition (SER) method that estimates emotions not previously defined in the SER model training. Conventional methods are limited to recognizing emotions defined by a single word. Moreover, we have the motivation to recognize unknown bipolar emotions such as ``I want to buy - I do not want to buy.'' In order to allow the model to define classes using sentences freely and to estimate unknown bipolar emotions, our proposed method expands upon the contrastive language-audio pre-training (CLAP) framework by introducing multi-class and multi-task settings. We also focus on purchase intention as a bipolar emotion and investigate the model's performance to zero-shot estimate it. This study is the first attempt to estimate purchase intention from speech directly. Experiments confirm that the results of zero-shot estimation by the proposed method are at the same level as those of the model trained by supervised learning., Comment: 5 pages, 3 figures, accepted for APSIPA 2024 ASC

Recent advancements in foundation models show promising capability in graphic design generation. Several studies have started employing Large Multimodal Models (LMMs) to evaluate graphic designs, assuming that LMMs can properly assess their quality, but it is unclear if the evaluation is reliable. One way to evaluate the quality of graphic design is to assess whether the design adheres to fundamental graphic design principles, which are the designer's common practice. In this paper, we compare the behavior of GPT-based evaluation and heuristic evaluation based on design principles using human annotations collected from 60 subjects. Our experiments reveal that, while GPTs cannot distinguish small details, they have a reasonably good correlation with human annotation and exhibit a similar tendency to heuristic metrics based on design principles, suggesting that they are indeed capable of assessing the quality of graphic design. Our dataset is available at https://cyberagentailab.github.io/Graphic-design-evaluation ., Comment: Accepted to SIGGRAPH Asia 2024 (Technical Communications Track)

This paper presents multimodal markup document models (MarkupDM) that can generate both markup language and images within interleaved multimodal documents. Unlike existing vision-and-language multimodal models, our MarkupDM tackles unique challenges critical to graphic design tasks: generating partial images that contribute to the overall appearance, often involving transparency and varying sizes, and understanding the syntax and semantics of markup languages, which play a fundamental role as a representational format of graphic designs. To address these challenges, we design an image quantizer to tokenize images of diverse sizes with transparency and modify a code language model to process markup languages and incorporate image modalities. We provide in-depth evaluations of our approach on three graphic design completion tasks: generating missing attribute values, images, and texts in graphic design templates. Results corroborate the effectiveness of our MarkupDM for graphic design tasks. We also discuss the strengths and weaknesses in detail, providing insights for future research on multimodal document generation., Comment: Project page: https://cyberagentailab.github.io/MarkupDM/

Insufficient recordings and the scarcity of anomalies present significant challenges in developing and validating robust anomaly detection systems for machine sounds. To address these limitations, we propose a novel approach for generating diverse anomalies in machine sound using a latent diffusion-based model that integrates an encoder-decoder framework. Our method utilizes the Flan-T5 model to encode captions derived from audio file metadata, enabling conditional generation through a carefully designed U-Net architecture. This approach aids our model in generating audio signals within the EnCodec latent space, ensuring high contextual relevance and quality. We objectively evaluated the quality of our generated sounds using the Fr\'echet Audio Distance (FAD) score and other metrics, demonstrating that our approach surpasses existing models in generating reliable machine audio that closely resembles actual abnormal conditions. The evaluation of the anomaly detection system using our generated data revealed a strong correlation, with the area under the curve (AUC) score differing by 4.8\% from the original, validating the effectiveness of our generated data. These results demonstrate the potential of our approach to enhance the evaluation and robustness of anomaly detection systems across varied and previously unseen conditions. Audio samples can be found at \url{https://hpworkhub.github.io/MIMII-Gen.github.io/}.

Due to scarcity of time-series data annotated with descriptive texts, training a model to generate descriptive texts for time-series data is challenging. In this study, we propose a method to systematically generate domain-independent descriptive texts from time-series data. We identify two distinct approaches for creating pairs of time-series data and descriptive texts: the forward approach and the backward approach. By implementing the novel backward approach, we create the Temporal Automated Captions for Observations (TACO) dataset. Experimental results demonstrate that a contrastive learning based model trained using the TACO dataset is capable of generating descriptive texts for time-series data in novel domains.

In the classical reach-avoid problem, autonomous mobile robots are tasked to reach a goal while avoiding obstacles. However, it is difficult to provide guarantees on the robot's performance when the obstacles form a narrow gap and the robot is a black-box (i.e. the dynamics are not known analytically, but interacting with the system is cheap). To address this challenge, this paper presents NeuralPARC. The method extends the authors' prior Piecewise Affine Reach-avoid Computation (PARC) method to systems modeled by rectified linear unit (ReLU) neural networks, which are trained to represent parameterized trajectory data demonstrated by the robot. NeuralPARC computes the reachable set of the network while accounting for modeling error, and returns a set of states and parameters with which the black-box system is guaranteed to reach the goal and avoid obstacles. Through numerical experiments, NeuralPARC is shown to outperform PARC in generating provably-safe extreme vehicle drift parking maneuvers, as well as enabling safety on an autonomous surface vehicle (ASV) subjected to large disturbances and controlled by a deep reinforcement learning (RL) policy., Comment: This work has been submitted for possible publication

We study the well-known coded caching problem in an online learning framework, wherein requests arrive sequentially, and an online policy can update the cache contents based on the history of requests seen thus far. We introduce a caching policy based on the Follow-The-Perturbed-Leader principle and show that for any time horizon T and any request sequence, it achieves a sub-linear regret of \mathcal{O}(\sqrt(T) ) with respect to an oracle that knows the request sequence beforehand. Our study marks the first examination of adversarial regret in the coded caching setup. Furthermore, we also address the issue of switching cost by establishing an upper bound on the expected number of cache updates made by our algorithm under unrestricted switching and also provide an upper bound on the regret under restricted switching when cache updates can only happen in a pre-specified subset of timeslots. Finally, we validate our theoretical insights with numerical results using a real-world dataset

The normalized maximum likelihood (NML) code length is widely used as a model selection criterion based on the minimum description length principle, where the model with the shortest NML code length is selected. A common method to calculate the NML code length is to use the sum (for a discrete model) or integral (for a continuous model) of a function defined by the distribution of the maximum likelihood estimator. While this method has been proven to correctly calculate the NML code length of discrete models, no proof has been provided for continuous cases. Consequently, it has remained unclear whether the method can accurately calculate the NML code length of continuous models. In this paper, we solve this problem affirmatively, proving that the method is also correct for continuous cases. Remarkably, completing the proof for continuous cases is non-trivial in that it cannot be achieved by merely replacing the sums in discrete cases with integrals, as the decomposition trick applied to sums in the discrete model case proof is not applicable to integrals in the continuous model case proof. To overcome this, we introduce a novel decomposition approach based on the coarea formula from geometric measure theory, which is essential to establishing our proof for continuous cases.

We propose and demonstrate the DisasterNeedFinder framework in order to provide appropriate information support for the Noto Peninsula Earthquake. In the event of a large-scale disaster, it is essential to accurately capture the ever-changing information needs. However, it is difficult to obtain appropriate information from the chaotic situation on the ground. Therefore, as a data-driven approach, we aim to pick up precise information needs at the site by integrally analyzing the location information of disaster victims and search information. It is difficult to make a clear estimation of information needs by just analyzing search history information in disaster areas, due to the large amount of noise and the small number of users. Therefore, the idea of assuming that the magnitude of information needs is not the volume of searches, but the degree of abnormalities in searches, enables an appropriate understanding of the information needs of the disaster victims. DNF has been continuously clarifying the information needs of disaster areas since the disaster strike, and has been recognized as a new approach to support disaster areas by being featured in the major Japanese media on several occasions.

Electromagnetic (EM) multipath interference is difficult to address with passive approaches due to two physical restrictions - the shared frequency of the initial and interfering signals and their variable incident angles. Thus, to address multipath interference, the spatial impedance must be adjusted in response to the incident angles of multiple signals with the same frequency, which is impossible with classic linear time-invariant (LTI) systems. We present a design concept for metasurface-based spatial filters to overcome LTI behavior and suppress multipath interference signals using a time-varying interlocking mechanism without any active biasing systems. The proposed devices are coupled to the first incoming wave to adjust the spatial impedance and suppress delayed waves in the time domain, which is validated numerically and experimentally. This study opens a new avenue for passive yet time-varying selective EM metasystems, enabling the adjustment of spatially complicated EM waves and fields even at the same frequency., Comment: 11 pages, 14 figures

This paper investigates the phenomenon of benign overfitting in binary classification problems with heavy-tailed input distributions, extending the analysis of maximum margin classifiers to $\alpha$ sub-exponential distributions ($\alpha \in (0, 2]$). This generalizes previous work focused on sub-gaussian inputs. We provide generalization error bounds for linear classifiers trained using gradient descent on unregularized logistic loss in this heavy-tailed setting. Our results show that, under certain conditions on the dimensionality $p$ and the distance between the centers of the distributions, the misclassification error of the maximum margin classifier asymptotically approaches the noise level, the theoretical optimal value. Moreover, we derive an upper bound on the learning rate $\beta$ for benign overfitting to occur and show that as the tail heaviness of the input distribution $\alpha$ increases, the upper bound on the learning rate decreases. These results demonstrate that benign overfitting persists even in settings with heavier-tailed inputs than previously studied, contributing to a deeper understanding of the phenomenon in more realistic data environments.

Quantum annealing has been recently studied to solve the shortest vector problem (SVP), where the norm of a lattice point vector is mapped to the problem Hamiltonian with the qudit encoding, Hamming-weight encoding, or binary encoding, and the problem to find the shortest vector is mapped to a problem to find a non-trivial first excited state. We here propose an alternative encoding and alternative quantum algorithm to solve the SVP: the one-hot encoding and the quantum imaginary-time algorithm with the folded spectrum (FS) method. We demonstrate that our approach is applicable to find the shortest vector with a variational quantum algorithm. The application of the FS method to the quantum annealing and simulated annealing is also discussed to solve the SVP. Our study shows wide potential applicability of the SVP in quantum computing frameworks., Comment: 9 pages, 9 figures

This paper presents the global analysis of two extended decreases of the galactic cosmic ray intensity observed by world-wide networks of ground-based detectors in 2012. This analysis is capable of separately deriving the cosmic ray density (or omnidirectional intensity) and anisotropy each as a function of time and rigidity. A simple diffusion model along the spiral field line between Earth and a cosmic-ray barrier indicates the long duration of these events resulting from about 190$^\circ$ eastern extension of a barrier such as an IP-shock followed by the sheath region and/or the corotating interaction region (CIR). It is suggested that the coronal mass ejection merging and compressing the preexisting CIR at its flank can produce such the extended barrier. The derived rigidity spectra of the density and anisotropy both vary in time during each event period. In particular we find that the temporal feature of the ``phantom Forbush decrease'' reported in an analyzed period is dependent on rigidity, looking quite different at different rigidities. From these rigidity spectra of the density and anisotropy, we derive the rigidity spectrum of the average parallel mean-free-path of pitch angle scattering along the spiral field line and infer the power spectrum of the magnetic fluctuation and its temporal variation. Possible physical cause of the strong rigidity dependence of the ``phantom Forbush decrease'' is also discussed. These results demonstrate the high-energy cosmic rays observed at Earth responding to remote space weather., Comment: Accepted for publication in ApJ

A compact energy-recovery linac (cERL) has been un-der construction at KEK since 2009 to develop key technologies for the energy-recovery linac. The cERL began operating in 2013 to create a high-current beam with a low-emittance beam with stable continuous wave (CW) superconducting cavities. Owing to the development of critical components, such as the DC gun, superconducting cavities, and the design of ideal beam transport optics, we have successfully established approximately 1 mA stable CW operation with a small beam emittance and extremely small beam loss. This study presents the details of our key technologies and experimental results for achieving 100% energy recovery operation with extremely small beam loss during a stable, approximately 1 mA CW beam operation., Comment: 11 pages, 16 figures

We propose a new series expansion method for the periapsis shift. The method formulates the periapsis shift in various spacetimes analytically without using special functions and provides simple and highly accurate approximate formulae. We derive new series representations for the periapsis shift in the Kerr and the Chazy-Curzon spacetimes by using the method, where the expansion parameter is defined as the eccentricity divided by the non-dimensional quantity that vanishes in the limit of the innermost stable circular orbit. That is to say, the expansion parameter denotes how eccentric the orbit is and how close it is to the innermost stable circular orbit. The smaller the eccentricity, the higher the accuracy of the formulae that are obtained by truncating the new series representations up to a finite number of terms. If the eccentricity is sufficiently small, the truncated new representations have higher accuracy than the post-Newtonian expansion formulae even in strong gravitational fields where the convergence of the post-Newtonian expansion formula is not guaranteed. On the other hand, even if the orbit is highly eccentric, the truncated new representations have comparable or higher accuracy than the post-Newtonian expansion formulae if the semi-major axis is sufficiently large. An exact formula for the periapsis shift of the quasi-circular orbit in the Chazy-Curzon spacetime is also obtained as a special case of the new series representation., Comment: 41 pages, 18 figures

We investigate compact stars in the noninteger power (NIP) model of $f(R)$ gravity theory, which includes the higher-curvature correction to the Einstein-Hilbert action. The mass-radius relation of the compact stars in the NIP model predicts large deviations from those in the general relativity in the low-mass region, potentially allowing us to test the NIP model by future astrophysical observations. We also study the nonvanishing scalar hair surrounding the compact star and demonstrate that the chameleon mechanism works efficiently. They result in distinct scalar profiles inside and outside the star, which implies screening the fifth force mediated by the scalar field., Comment: 35 pages, 27 figures