Search

Large Language Models (LLMs) have demonstrated a number of human-like abilities, however the empathic understanding and emotional state of LLMs is yet to be aligned to that of humans. In this work, we investigate how the emotional state of generative LLM agents evolves as they perceive new events, introducing a novel architecture in which new experiences are compared to past memories. Through this comparison, the agent gains the ability to understand new experiences in context, which according to the appraisal theory of emotion is vital in emotion creation. First, the agent perceives new experiences as time series text data. After perceiving each new input, the agent generates a summary of past relevant memories, referred to as the norm, and compares the new experience to this norm. Through this comparison we can analyse how the agent reacts to the new experience in context. The PANAS, a test of affect, is administered to the agent, capturing the emotional state of the agent after the perception of the new event. Finally, the new experience is then added to the agents memory to be used in the creation of future norms. By creating multiple experiences in natural language from emotionally charged situations, we test the proposed architecture on a wide range of scenarios. The mixed results suggests that introducing context can occasionally improve the emotional alignment of the agent, but further study and comparison with human evaluators is necessary. We hope that this paper is another step towards the alignment of generative agents., Comment: 14 pages, 6 figures

Traditional approaches in offline reinforcement learning aim to learn the optimal policy that maximizes the cumulative reward, also known as return. However, as applications broaden, it becomes increasingly crucial to train agents that not only maximize the returns, but align the actual return with a specified target return, giving control over the agent's performance. Decision Transformer (DT) optimizes a policy that generates actions conditioned on the target return through supervised learning and is equipped with a mechanism to control the agent using the target return. However, the action generation is hardly influenced by the target return because DT's self-attention allocates scarce attention scores to the return tokens. In this paper, we propose Return-Aligned Decision Transformer (RADT), designed to effectively align the actual return with the target return. RADT utilizes features extracted by paying attention solely to the return, enabling the action generation to consistently depend on the target return. Extensive experiments show that RADT reduces the discrepancies between the actual return and the target return of DT-based methods.

This paper proposes a compression framework for adjacency matrices of weighted graphs based on graph filter banks. Adjacency matrices are widely used mathematical representations of graphs and are used in various applications in signal processing, machine learning, and data mining. In many problems of interest, these adjacency matrices can be large, so efficient compression methods are crucial. In this paper, we propose a lossy compression of weighted adjacency matrices, where the binary adjacency information is encoded losslessly (so the topological information of the graph is preserved) while the edge weights are compressed lossily. For the edge weight compression, the target graph is converted into a line graph, whose nodes correspond to the edges of the original graph, and where the original edge weights are regarded as a graph signal on the line graph. We then transform the edge weights on the line graph with a graph filter bank for sparse representation. Experiments on synthetic data validate the effectiveness of the proposed method by comparing it with existing lossy matrix compression methods.

Explainable artificial intelligence (XAI) has helped elucidate the internal mechanisms of machine learning algorithms, bolstering their reliability by demonstrating the basis of their predictions. Several XAI models consider causal relationships to explain models by examining the input-output relationships of prediction models and the dependencies between features. The majority of these models have been based their explanations on counterfactual probabilities, assuming that the causal graph is known. However, this assumption complicates the application of such models to real data, given that the causal relationships between features are unknown in most cases. Thus, this study proposed a novel XAI framework that relaxed the constraint that the causal graph is known. This framework leveraged counterfactual probabilities and additional prior information on causal structure, facilitating the integration of a causal graph estimated through causal discovery methods and a black-box classification model. Furthermore, explanatory scores were estimated based on counterfactual probabilities. Numerical experiments conducted employing artificial data confirmed the possibility of estimating the explanatory score more accurately than in the absence of a causal graph. Finally, as an application to real data, we constructed a classification model of credit ratings assigned by Shiga Bank, Shiga prefecture, Japan. We demonstrated the effectiveness of the proposed method in cases where the causal graph is unknown.

Recent studies have shown that the magnetic properties of the ferromagnetic perovskite oxide La$_{2/3}$ Sr$_{1/3}$ MnO$_3$ (LSMO) grown on an SrTiO3 (STO) substrate, such as its magnetic moment and Curie temperature, can be improved by releasing the film from the substrate. However, the microscopic origin of this enhancement is not yet well understood. In this study, we use synchrotron radiation measurements to investigate the mechanism of ferromagnetism enhancement in an LSMO membrane released from an STO substrate by dissolving a water-soluble Sr$_4$Al$_2$O$_7$ buffer layer. Using resonant photoemission spectroscopy on the as-grown LSMO film and LSMO membrane, we elucidate that the strain release from the STO substrate enhances the itineracy of the Mn-3d electrons via p-d hybridization, and this strengthens the double-exchange interaction. The reinforcement of the double-exchange interaction, in turn, improves the ferromagnetism of LSMO.

Doping of liquid argon TPCs (LArTPCs) with a small concentration of xenon is a technique for light-shifting and facilitates the detection of the liquid argon scintillation light. In this paper, we present the results of the first doping test ever performed in a kiloton-scale LArTPC. From February to May 2020, we carried out this special run in the single-phase DUNE Far Detector prototype (ProtoDUNE-SP) at CERN, featuring 720 t of total liquid argon mass with 410 t of fiducial mass. A 5.4 ppm nitrogen contamination was present during the xenon doping campaign. The goal of the run was to measure the light and charge response of the detector to the addition of xenon, up to a concentration of 18.8 ppm. The main purpose was to test the possibility for reduction of non-uniformities in light collection, caused by deployment of photon detectors only within the anode planes. Light collection was analysed as a function of the xenon concentration, by using the pre-existing photon detection system (PDS) of ProtoDUNE-SP and an additional smaller set-up installed specifically for this run. In this paper we first summarize our current understanding of the argon-xenon energy transfer process and the impact of the presence of nitrogen in argon with and without xenon dopant. We then describe the key elements of ProtoDUNE-SP and the injection method deployed. Two dedicated photon detectors were able to collect the light produced by xenon and the total light. The ratio of these components was measured to be about 0.65 as 18.8 ppm of xenon were injected. We performed studies of the collection efficiency as a function of the distance between tracks and light detectors, demonstrating enhanced uniformity of response for the anode-mounted PDS. We also show that xenon doping can substantially recover light losses due to contamination of the liquid argon by nitrogen., Comment: 36 pages, 20 figures. Corrected author list; corrected typos across paper and polished text

We prove the solvability and nilpotency of Kac--Paljutkin's finite quantum group and Sekine quantum groups and we classify the solvable series of Kac--Paljutkin's finite quantum group via Cohen--Westreich's Burnside theorem. Some semisimple quasitriangular Hopf algebras of dimensions $2pq$ are also studied. In Appendix A, we give a direct computation of the universal $R$-matrices for Kac--Paljutkin's $8$-dimensional finite quantum group., Comment: 47 pages

Majorana fermions are spin-1/2 neutral particles that are their own antiparticles and were initially predicted by Ettore Majorana in particle physics but their observation still remains elusive. The concept of Majorana fermions has been borrowed into condensed matter physics, where, unlike particle physics, Majorana fermions emerge as zero-energy quasiparticles that can be engineered by combining electrons and holes and have therefore been coined Majorana zero modes. In this review, we provide a pedagogical explanation of the basic properties of Majorana zero modes in unconventional superconductors and their consequences in experimental observables, putting a special emphasis on the initial theoretical discoveries. In particular, we first show that Majorana zero modes are self-conjugated and emerge as a special type of zero energy surface Andreev bound states at the boundary of unconventional superconductors. We then explore Majorana zero modes in one-dimensional spin-polarized $p$-wave superconductors, where we address the formation of topological superconductivity and the physical realization in superconductor-semiconductor hybrids. In this part we highlight that Majorana quasiparticles appear as zero-energy edge states, exhibiting charge neutrality, spin-polarized, and spatial nonlocality as unique properties that can be already seen from their energies and wavefunctions. Next, we discuss analytically obtained Green's functions of $p$-wave superconductors and demonstrate that the emergence of Majorana zero modes is always accompanied by the formation of odd-frequency spin-triplet pairing as a unique result of the self-conjugate nature of Majorana zero modes. We finally address the signatures of Majorana zero modes in tunneling spectroscopy, including the anomalous proximity effect, and the phase-biased Josephson effect.

We report on the development of a novel pixel charge readout system, Grid Activated Multi-scale pixel readout (GAMPix), which is under development for use in the GammaTPC gamma ray instrument concept. GammaTPC is being developed to optimize the use of liquid argon time projection chamber technology for gamma ray astrophysics, for which a fine grained low power charge readout is essential. GAMPix uses a new architecture with coarse and fine scale instrumented electrodes to solve the twin problems of loss of measured charge after diffusion, and high readout power. Fundamentally, it enables low noise and ultra low power charge readout at the spatial scale limited by diffusion in a time projection chamber, and has other possibly applications, including future DUNE modules., Comment: 23 pages, 18 figures

We demonstrate that quantum error correction is realized by the renormalization group in scalar field theories. We construct $q$-level states by using coherent states in the IR region. By acting on them the inverse of the unitary operator $U$ that describes the renormalization group flow of the ground state, we encode them into states in the UV region. We find that the Knill-Laflamme condition is satisfied for operators that create coherent states. We verify this to the first order in the perturbation theory. This result suggests a general relationship between the renormalization group and quantum error correction and should give insights into understanding a role played by them in the gauge/gravity correspondence., Comment: 21 pages

We present a combined growth and transport study of superconducting single-crystalline NdO$_{0.5}$F$_{0.5}$BiS$_2$. Evidence of two-dimensional superconductivity with significant phase fluctuations of preformed Cooper pairs preceding the superconducting transition is reported. This result is based on three key observations. (1) The resistive superconducting transition temperature $T_c$ (defined by resistivity $\rho \rightarrow 0$) increases with increasing disorder. (2) As $T\rightarrow T_c$, the conductivity diverges significantly faster than what is expected from Gaussian fluctuations in two and three dimensions. (3) Non-Ohmic resistance behavior is observed in the superconducting state. Altogether, our observations are consistent with a temperature regime of phase-fluctuating superconductivity. The crystal structure with magnetic ordering tendencies in the NdO$_{0.5}$F$_{0.5}$ layers and (super)conductivity in the BiS$_2$ layers is likely responsible for the two-dimensional phase fluctuations. As such, NdO$_{0.5}$F$_{0.5}$BiS$_2$ falls into the class of unconventional ``laminar" bulk superconductors that include cuprate materials and 4Hb-TaS$_2$.

Dynamical behaviors of complex interacting systems, including brain activities, financial price movements, and physical collective phenomena, are associated with underlying interactions between the system's components. The issue of uncovering interaction relations in such systems using observable dynamics is called relational inference. In this study, we propose a Diffusion model for Relational Inference (DiffRI), inspired by a self-supervised method for probabilistic time series imputation. DiffRI learns to infer the probability of the presence of connections between components through conditional diffusion modeling.

The deep neural networks are known to be vulnerable to well-designed adversarial attacks. The most successful defense technique based on adversarial training (AT) can achieve optimal robustness against particular attacks but cannot generalize well to unseen attacks. Another effective defense technique based on adversarial purification (AP) can enhance generalization but cannot achieve optimal robustness. Meanwhile, both methods share one common limitation on the degraded standard accuracy. To mitigate these issues, we propose a novel pipeline to acquire the robust purifier model, named Adversarial Training on Purification (AToP), which comprises two components: perturbation destruction by random transforms (RT) and purifier model fine-tuned (FT) by adversarial loss. RT is essential to avoid overlearning to known attacks, resulting in the robustness generalization to unseen attacks, and FT is essential for the improvement of robustness. To evaluate our method in an efficient and scalable way, we conduct extensive experiments on CIFAR-10, CIFAR-100, and ImageNette to demonstrate that our method achieves optimal robustness and exhibits generalization ability against unseen attacks.

SS 433 is a microquasar, a stellar binary system with collimated relativistic jets. We observed SS 433 in gamma rays using the High Energy Stereoscopic System (H.E.S.S.), finding an energy-dependent shift in the apparent position of the gamma-ray emission of the parsec-scale jets. These observations trace the energetic electron population and indicate the gamma rays are produced by inverse-Compton scattering. Modelling of the energy-dependent gamma-ray morphology constrains the location of particle acceleration and requires an abrupt deceleration of the jet flow. We infer the presence of shocks on either side of the binary system at distances of 25 to 30 parsecs and conclude that self-collimation of the precessing jets forms the shocks, which then efficiently accelerate electrons., Comment: Submitted 20th Apr. 2023, published 25th January 2024 (accepted version)

Dialogue systems controlled by predefined or rule-based scenarios derived from counseling techniques, such as cognitive behavioral therapy (CBT), play an important role in mental health apps. Despite the need for responsible responses, it is conceivable that using the newly emerging LLMs to generate contextually relevant utterances will enhance these apps. In this study, we construct dialogue modules based on a CBT scenario focused on conventional Socratic questioning using two kinds of LLMs: a Transformer-based dialogue model further trained with a social media empathetic counseling dataset, provided by Osaka Prefecture (OsakaED), and GPT-4, a state-of-the art LLM created by OpenAI. By comparing systems that use LLM-generated responses with those that do not, we investigate the impact of generated responses on subjective evaluations such as mood change, cognitive change, and dialogue quality (e.g., empathy). As a result, no notable improvements are observed when using the OsakaED model. When using GPT-4, the amount of mood change, empathy, and other dialogue qualities improve significantly. Results suggest that GPT-4 possesses a high counseling ability. However, they also indicate that even when using a dialogue model trained with a human counseling dataset, it does not necessarily yield better outcomes compared to scenario-based dialogues. While presenting LLM-generated responses, including GPT-4, and having them interact directly with users in real-life mental health care services may raise ethical issues, it is still possible for human professionals to produce example responses or response templates using LLMs in advance in systems that use rules, scenarios, or example responses., Comment: Accepted by IWSDS2024

State estimation for legged robots is challenging due to their highly dynamic motion and limitations imposed by sensor accuracy. By integrating Kalman filtering, optimization, and learning-based modalities, we propose a hybrid solution that combines proprioception and exteroceptive information for estimating the state of the robot's trunk. Leveraging joint encoder and IMU measurements, our Kalman filter is enhanced through a single-rigid body model that incorporates ground reaction force control outputs from convex Model Predictive Control optimization. The estimation is further refined through Gated Recurrent Units, which also considers semantic insights and robot height from a Vision Transformer autoencoder applied on depth images. This framework not only furnishes accurate robot state estimates, including uncertainty evaluations, but can minimize the nonlinear errors that arise from sensor measurements and model simplifications through learning. The proposed methodology is evaluated in hardware using a quadruped robot on various terrains, yielding a 65% improvement on the Root Mean Squared Error compared to our VIO SLAM baseline. Code example: https://github.com/AlexS28/OptiState, Comment: Accepted to the 2024 IEEE International Conference on Robotics and Automation (ICRA), May 13-17, in Yokohama, Japan. 7 pages, 5 figures, 1 table

Human activities generate various event sequences such as taxi trip records, bike-sharing pick-ups, crime occurrence, and infectious disease transmission. The point process is widely used in many applications to predict such events related to human activities. However, point processes present two problems in predicting events related to human activities. First, recent high-performance point process models require the input of sufficient numbers of events collected over a long period (i.e., long sequences) for training, which are often unavailable in realistic situations. Second, the long-term predictions required in real-world applications are difficult. To tackle these problems, we propose a novel meta-learning approach for periodicity-aware prediction of future events given short sequences. The proposed method first embeds short sequences into hidden representations (i.e., task representations) via recurrent neural networks for creating predictions from short sequences. It then models the intensity of the point process by monotonic neural networks (MNNs), with the input being the task representations. We transfer the prior knowledge learned from related tasks and can improve event prediction given short sequences of target tasks. We design the MNNs to explicitly take temporal periodic patterns into account, contributing to improved long-term prediction performance. Experiments on multiple real-world datasets demonstrate that the proposed method has higher prediction performance than existing alternatives.

We perform a numerical simulation of compression of a highly porous dust aggregate of monodisperse spheres. We find that the average interparticle normal force within the aggregate is inversely proportional to both the filling factor and the average coordination number, and we also derive this relation theoretically. Our findings would be applicable for granular matter of arbitrary structures, as long as the constituent particles are monodisperse spheres., Comment: 9 pages, 7 figures. Accepted for publication in PRE

The polynomial machine learning potentials (MLPs) described by polynomial rotational invariants have been systematically developed for various systems and used in diverse applications in crystalline states. In this study, we systematically investigate the predictive power of the polynomial MLPs for liquid structural properties in 22 elemental systems with diverse chemical bonding properties, including those showing anomalous melting behavior, such as Si, Ge, and Bi. We compare liquid structural properties obtained from molecular dynamics simulations using the density functional theory (DFT) calculation, the polynomial MLPs, and other interatomic potentials in the literature. The current results demonstrate that the polynomial MLPs consistently exhibit high predictive power for liquid structural properties with the same accuracy as that of typical DFT calculations., Comment: REVTeX 4-2, 30 pages with 35 figures

We present the luminosity functions and stellar mass functions of supernova (SN) host galaxies and test if they differ from the functions of normal field galaxies. We utilize homogeneous samples consisting of 273 SNe Ia ($z\leq0.3$) and 44 core-collapse (CC) SNe ($z \leq 0.1$) from the Sloan Digital Sky Survey (SDSS) II Supernova Survey and the high-signal-to-noise-ratio photometry of galaxies from the Hyper Suprime-Cam Subaru Strategic Program (HSC SSP). SN hosts are classified into star-forming and passive galaxy groups based on the spectral energy distribution (SED) fitting. We find that the SN host luminosity functions and stellar mass functions deviate from those of normal field galaxies. Star-forming galaxies dominate the low-mass end of the SN Ia host mass function, while passive galaxies dominate the high-mass end. CC SNe are predominantly hosted by star-forming galaxies. In addition, intermediate-mass hosts produce CC SNe with the highest efficiency, while the efficiency of producing SNe Ia monotonically increases as the hosts become more massive. Furthermore, We derive the pseudo mass normalized SN rates (pSNuM) based on the mass functions. We find that the star-forming component of pSNuM$_{Ia}$ is less sensitive to the changes in stellar mass, in comparison with the total rate. The behavior of pSNuM$_{CC}$ suggests that the CC rate is proportional to the star-forming rate.

We study the problem of completing various visual document understanding (VDU) tasks, e.g., question answering and information extraction, on real-world documents through human-written instructions. To this end, we propose InstructDoc, the first large-scale collection of 30 publicly available VDU datasets, each with diverse instructions in a unified format, which covers a wide range of 12 tasks and includes open document types/formats. Furthermore, to enhance the generalization performance on VDU tasks, we design a new instruction-based document reading and understanding model, InstructDr, that connects document images, image encoders, and large language models (LLMs) through a trainable bridging module. Experiments demonstrate that InstructDr can effectively adapt to new VDU datasets, tasks, and domains via given instructions and outperforms existing multimodal LLMs and ChatGPT without specific training., Comment: Accepted by AAAI2024; project page: https://github.com/nttmdlab-nlp/InstructDoc

This paper proposes a level set-based method for optimizing shell structures with large design changes in shape and topology. Conventional shell optimization methods, whether parametric or nonparametric, often only allow limited design changes in shape. In the proposed method, the shell structure is defined as the isosurface of a level set function. The level set function is iteratively updated based on the shape sensitivity on the surface mesh. Therefore, the proposed method can represent an arbitrary manifold surface while dealing with topological changes, for example, from a spherical surface to a toroidal surface. We applied the proposed method to the mean compliance minimization problems of 3D shell structural designs for dome, bending plate and cantilever beam examples to demonstrate its efficacy of the proposed method., Comment: 13 pages, 13 figures

Our knowledge of relations between supermassive black holes and their host galaxies at $z\gtrsim1$ is still limited, even though being actively sought out to $z\sim6$. Here, we use the high resolution and sensitivity of JWST to measure the host galaxy properties for 61 X-ray-selected type-I AGNs at $0.7

The SliT ASIC is a readout chip for the silicon strip detector to be used at the J-PARC muon $g-2$/EDM experiment. The production version of SliT128D was designed and mass production was finished. A quality assurance test method for bare SliT128D chips was developed to provide a sufficient number of chips for the experiment. The quality assurance test of the SliT128D chips was performed and 5735 chips were inspected. No defect was observed in chips of 84.3%. Accepting a few channels with poor time walk performance out of 128 channels per chip, more than 90% yield can be achieved, which is sufficient to construct the whole detector., Comment: 5 pages, 8 figures

While various methods have been proposed to disentangle the progenitor system for Type Ia supernovae, their origin is still unclear. Circumstellar environment is a key to distinguishing between the double-degenerate (DD) and single-degenerate (SD) scenarios since a dense wind cavity is expected only in the case of the SD system. We perform spatially resolved X-ray spectroscopy of Tycho's supernova remnant (SNR) with XMM-Newton and reveal the three-dimensional velocity structure of the expanding shock-heated ejecta measured from Doppler-broadened lines of intermediate-mass elements. Obtained velocity profiles are fairly consistent with those expected from a uniformly expanding ejecta model near the center, whereas we discover a rapid deceleration ($\sim4000$ km s$^{-1}$ to $\sim1000$ km s$^{-1}$) near the edge of the remnant in almost every direction. The result strongly supports the presence of a dense wall entirely surrounding the remnant, which is confirmed also by our hydrodynamical simulation. We thus conclude that Tycho's SNR is likely of the SD origin. Our new method will be useful for understanding progenitor systems of Type Ia SNRs in the era of high-angular/energy resolution X-ray astronomy with microcalorimeters., Comment: 15 pages, 12 figures, accepted for publication in ApJ

We study the interplay between the Kondo and superconducting (SC) proximity effects, taking place in a triangular triple quantum dot (TTQD) connected to one normal and two SC leads. This system shows various quantum phases. Without the SC leads, the lowest two states that belong to the different spin sectors, $S=0$ and $S=1/2$, become energetically very close to each other near half filling. The singlet one is a Kondo-screened state by conduction electrons from the normal lead, and the doublet one is a resonating valence bond state with unpaired free spin which remains unscreened. Furthermore, when one additional electron enters the TTQD, the ground state becomes a doublet in which the $S=1$ local moment due to the Nagaoka ferromagnetism is partially screened by conduction electrons.The Cooper pairs penetrating into the TTQD from the SC leads reconstruct the wavefunctions and vary the phase boundaries between these quantum states in the parameter space. We calculate ground-state phase diagrams using the numerical renormalization group, and show that the SC proximity effect induces a reentrant transition in-between the three- and four-electron fillings., Comment: 7 pages, 3 figures

We study the nonlocal magnetotransport through a strongly correlated quantum dot, connected to multiple terminals consisting of two normal and one superconducting (SC) leads. Specifically, we present a comprehensive view on the interplay between the crossed Andreev reflection (CAR), the Kondo effect, and the Zeeman splitting at zero temperature in the large SC gap limit. The ground state of this network shows an interesting variety, which varies continuously with the system parameters, such as the coupling strength $\Gamma_S^{}$ between the SC lead and the quantum dot, the Coulomb repulsion $U$, the impurity level $\varepsilon_d^{}$, and the magnetic field $b$. We show, using the many-body optical theorem which is derived from the Fermi-liquid theory, that the nonlocal conductance is determined by the transmission rate of the Cooper pairs $\mathcal{T}_{\mathrm{CP}}^{} = \frac{1}{4} \sin^2 \Theta\, \sin^2 \bigl(\delta_{\uparrow}+ \delta_{\downarrow})$ and that of the Bogoliubov particles $\mathcal{T}_{\mathrm{BG}}^{}= \frac{1}{2}\sum_{\sigma} \sin^2 \delta_{\sigma}^{}$. Here, $\delta_\sigma^{}$ is the phase shift of the renormalized Bogoliubov particles, and $\Theta \equiv \cot^{-1} (\xi_d^{}/ \Gamma_S^{})$ is the Bogoliubov-rotation angle in the Nambu pseudo spin space, with $\xi_d^{} =\varepsilon_d^{}+U/2$. It is also demonstrated, using Wilson's numerical renormalization group approach, that the CAR is enhanced in the crossover region between the Kondo regime and the SC-proximity-dominated regime at zero magnetic field. The magnetic fields induce another crossover between the Zeeman-dominated regime and the SC-dominated regime. We find that the CAR is enhanced and becomes less sensitive to magnetic fields in the SC-dominated regime close to the crossover region spreading over the angular range of $\pi/4 \lesssim \Theta \lesssim 3\pi/4$., Comment: 21 pages, 12 figures

The increasing demand for intelligent systems capable of interpreting and reasoning about visual content requires the development of large Vision-and-Language Models (VLMs) that are not only accurate but also have explicit reasoning capabilities. This paper presents a novel approach to develop a VLM with the ability to conduct explicit reasoning based on visual content and textual instructions. We introduce a system that can ask a question to acquire necessary knowledge, thereby enhancing the robustness and explicability of the reasoning process. To this end, we developed a novel dataset generated by a Large Language Model (LLM), designed to promote chain-of-thought reasoning combined with a question-asking mechanism. The dataset covers a range of tasks, from common ones like caption generation to specialized VQA tasks that require expert knowledge. Furthermore, using the dataset we created, we fine-tuned an existing VLM. This training enabled the models to generate questions and perform iterative reasoning during inference. The results demonstrated a stride toward a more robust, accurate, and interpretable VLM, capable of reasoning explicitly and seeking information proactively when confronted with ambiguous visual input.

The global drive towards carbon neutrality has led to a significant increase in the number of power plants based on renewable energy sources (RES). Concurrently, numerous households are adopting RES to generate their own energy, aiming to decrease both electricity costs and carbon footprints. To support these users, many papers have been devoted to developing optimal investment strategies for residential energy systems. However, there is still a significant gap as these studies often neglect important aspects like carbon neutrality. For this reason, in this paper, we explore the concept of net-zero energy houses (ZEHs) -- houses designed to have an annual net energy consumption around zero -- by presenting a constrained optimization problem to find the optimal number of photovoltaic panels and the optimal size of the battery system for home integration. Solving this constrained optimization problem is difficult due to its nonconvex constraints. Nevertheless, by applying a series of transformations, we reveal that it is possible to find an equivalent linear programming (LP) problem which is computationally tractable. The attainment of ZEH can be tackled by introducing a single constraint in the optimization problem. Additionally, we propose a sharing economy approach to the investment problem, offering a strategy that could potentially reduce investment costs and facilitate the attainment of ZEH more efficiently. Finally, we apply the proposed frameworks to a neighborhood in Japan as a case study, demonstrating the potential for long-term ZEH attainment. The results show that, under the right incentive, users can achieve ZEH, reduce their electricity costs and have a minimal impact on the main grid., Comment: 13 pages, 13 figures, 5 tables

We study two-dimensional $s$-wave topological superconductivity with Rashba spin-orbit coupling and Zeeman field in Penrose and Ammann-Beenker quasicrystals. By solving the Bogoliubov-de Gennes equations self-consistently for not only the superconducting order parameter, but also the spin-dependent Hartree potential, we show the stable occurrence of TSC with broken time-reversal symmetry in both Penrose and Ammann-Beenker quasicrystals. The topological nature of the quasicrystalline system is signified by the Bott index $B$. Topological phase transitions are found to occur, where $B$ changes between 0 and $\pm 1$, as the chemical potential or Zeeman field is varied. In terms of self-consistent solutions, we demonstrate the existence of a Majorana zero mode per edge or vortex when $B=\pm 1$, consistently with the bulk-edge/defect correspondence for periodic systems., Comment: 11 pages, 10 figures

With large training datasets and massive amounts of computing sources, large language models (LLMs) achieve remarkable performance in comprehensive and generative ability. Based on those powerful LLMs, the model fine-tuned with domain-specific datasets posseses more specialized knowledge and thus is more practical like medical LLMs. However, the existing fine-tuned medical LLMs are limited to general medical knowledge with English language. For disease-specific problems, the model's response is inaccurate and sometimes even completely irrelevant, especially when using a language other than English. In this work, we focus on the particular disease of Epilepsy with Japanese language and introduce a customized LLM termed as EpilepsyLLM. Our model is trained from the pre-trained LLM by fine-tuning technique using datasets from the epilepsy domain. The datasets contain knowledge of basic information about disease, common treatment methods and drugs, and important notes in life and work. The experimental results demonstrate that EpilepsyLLM can provide more reliable and specialized medical knowledge responses.

We present photometric, spectroscopic and polarimetric observations of the intermediate-luminosity Type IIP supernova (SN) 2021gmj from 1 to 386 days after the explosion. The peak absolute V-band magnitude of SN 2021gmj is -15.5 mag, which is fainter than that of normal Type IIP SNe. The spectral evolution of SN 2021gmj resembles that of other sub-luminous supernovae: the optical spectra show narrow P-Cygni profiles, indicating a low expansion velocity. We estimate the progenitor mass to be about 12 Msun from the nebular spectrum and the 56Ni mass to be about 0.02 Msun from the bolometric light curve. We also derive the explosion energy to be about 3 x 10^{50} erg by comparing numerical light curve models with the observed light curves. Polarization in the plateau phase is not very large, suggesting nearly spherical outer envelope. The early photometric observations capture the rapid rise of the light curve, which is likely due to the interaction with a circumstellar material (CSM). The broad emission feature formed by highly-ionized lines on top of a blue continuum in the earliest spectrum gives further indication of the CSM at the vicinity of the progenitor. Our work suggests that a relatively low-mass progenitor of an intermediate-luminosity Type IIP SN can also experience an enhanced mass loss just before the explosion, as suggested for normal Type IIP SNe., Comment: 18 pages, 16 figures, resubmitted to MNRAS after addressing referee comments

In conventional ionization chambers (ICs) using P-10 (Ar+CH4) gas, as the atomic number (Z) of the ion beams increases in the energy region of 200-300 MeV/u, the Z resolution deteriorates rapidly when Z>70. This degradation is attributed to substantial energy loss straggling caused by charge state fluctuation when the beams traverse a gas medium. The energy loss straggling increases when the beams cannot attain charge state equilibrium in the IC gas. In this study, a xenon-based gas (Xe+CH4), exhibiting a sufficiently large charge state changing cross section, was used in the IC to reach charge state equilibrium. The responses of ICs with P-10 and the xenon-based gases were examined using 238U beams and cocktail radioactive isotope (RI) beams with Z=40-90 at the RI Beam Factory (RIBF). For 238U beams at 165-344 MeV/u, the P-10 gas IC yielded an energy resolution of 1.9-3.0% in full width at half maximum (FWHM), which proved inadequate for Z identification in the uranium region. In contrast, the xenon-based gas IC demonstrated a satisfactory energy resolution of 1.4-1.6%. When using cocktail RI beams, a Z resolution of 1.28 and 0.74 was achieved by the P-10 and the xenon-based gas ICs, respectively, for beams with Z=84-88 at 200 MeV/u. The contrast in Z resolutions between the P-10 and the xenon-based gas ICs was effectively elucidated by the energy loss straggling model, incorporating collisional straggling and straggling due to charge state changes in the IC gases. The xenon-based gas IC, with more than 3sigma Z separation across a broad Z range (Z=40-90), emerged as a practical solution for Z identification of heavy ion beams., Comment: 24 pages, 13 figures, 1 table

We have been developing silicon-on-insulator (SOI) pixel detectors with a pinned depleted diode (PDD) structure, named "XRPIX", for X-ray astronomy. The PDD structure is formed in a thick p-type substrate, to which high negative voltage is applied to make it fully depleted. A pinned p-well is introduced at the backside of the insulator layer to reduce a dark current generation at the Si-SiO$_{2}$ interface and to fix the back-gate voltage of the SOI transistors. An n-well is further introduced between the p-well and the substrate to make a potential barrier between them and suppress a leakage current. An optimization study on the n-well dopant concentration is necessary because a higher dopant concentration could result in a higher potential barrier but also in a larger sense-node capacitance leading to a lower spectroscopic performance, and vice versa. Based on a device simulation, we fabricated five candidate chips having different n-well dopant concentrations. We successfully found out the best n-well design, which suppressed a large leakage current and showed satisfactory X-ray spectroscopic performance. Too low and too high n-well dopant concentration chips showed a large leakage current and degraded X-ray spectroscopic performance, respectively. We also found that the dependency of X-ray spectroscopic performance on the n-well dopant concentration can be largely explained by the difference in sense-node capacitance., Comment: 9 pages, 9 figures, accepted for publication in NIM A

The ratio of branching fractions $R(D^{*}) = \mathcal{B}(\overline{B} \rightarrow D^{*} \tau^{-} \overline{\nu}_{\tau})$/$\mathcal{B} (\overline{B} \rightarrow D^{*} \ell^{-} \overline{\nu}_{\ell})$, where $\ell$ is an electron or muon, is measured using a Belle~II data sample with an integrated luminosity of $189~\mathrm{fb}^{-1}$ at the SuperKEKB asymmetric-energy $e^{+} e^{-}$ collider. Data is collected at the $\Upsilon(\mathrm{4S})$ resonance, and one $B$ meson in the $\Upsilon(\mathrm{4S})\rightarrow B\overline{B}$ decay is fully reconstructed in hadronic decay modes. The accompanying signal $B$ meson is reconstructed as $\overline{B}\rightarrow D^{*} \tau^{-}\overline{\nu}_{\tau}$ using leptonic $\tau$ decays. The normalization decay, $\overline{B}\rightarrow D^{*} \ell^{-} \overline{\nu}_{\ell}$, where $\ell$ is an electron or muon, produces the same observable final state particles. The ratio of branching fractions is extracted in a simultaneous fit to two signal-discriminating variables in both channels and yields $R(D^{*}) = 0.262~_{-0.039}^{+0.041}(\mathrm{stat})~_{-0.032}^{+0.035}(\mathrm{syst})$. This result is consistent with the current world average and with standard model predictions., Comment: 16 pages, 17 figures, submitted to PRD

We present a novel $\mathcal{N} = 2 $ $\mathbb{Z}_2^2$-graded supersymmetric quantum mechanics ($\mathbb{Z}_2^2$-SQM) which has different features from those introduced so far. It is a two-dimensional (two-particle) system and is the first example of the quantum mechanical realization of an eight-dimensional irrep of the $\mathcal{N}=2$ $\mathbb{Z}_2^2$-supersymmetry algebra. The $\mathbb{Z}_2^2$-SQM is obtained by quantizing the one-dimensional classical system derived by dimensional reduction from the two-dimensional $\mathbb{Z}_2^2$-supersymmetric Lagrangian of $\mathcal{N}=1$, which we constructed in our previous work. The ground states of the $\mathbb{Z}_2^2$-SQM are also investigated., Comment: 16 pages, no figure

Molecular lines are powerful diagnostics of the physical and chemical properties of the interstellar medium (ISM). These ISM properties, which affect future star formation, are expected to differ in starburst galaxies from those of more quiescent galaxies. We investigate the ISM properties in the central molecular zone of the nearby starburst galaxy NGC 253 using the ultra-wide millimeter spectral scan survey from the ALMA Large Program ALCHEMI. We present an atlas of velocity-integrated images at a 1".6 resolution of 148 unblended transitions from 44 species, including the first extragalactic detection of HCNH$^+$ and the first interferometric images of C$_3$H$^+$, NO, HCS$^+$. We conduct a principal component analysis (PCA) on these images to extract correlated chemical species and to identify key groups of diagnostic transitions. To the best of our knowledge, our dataset is currently the largest astronomical set of molecular lines to which PCA has been applied. The PCA can categorize transitions coming from different physical components in NGC 253 such as i) young starburst tracers characterized by high-excitation transitions of HC$_3$N and complex organic molecules (COMs) versus tracers of on-going star formation (radio recombination lines) and high-excitation transitions of CCH and CN tracing PDRs, ii) tracers of cloud-collision-induced shocks (low-excitation transitions of CH$_3$OH, HNCO, HOCO$^+$, and OCS) versus shocks from star-formation-induced outflows (high-excitation transitions of SiO), as well as iii) outflows showing emission from HOC$^+$, CCH, H$_3$O$^+$, CO isotopologues, HCN, HCO$^+$, CS, and CN. Our findings show these intensities vary with galactic dynamics, star formation activities, and stellar feedback., Comment: 65 pages, 39 figures. Accepted for publication in ApJS

We present the final results of our search for new Milky Way (MW) satellites using the data from the Hyper Suprime-Cam (HSC) Subaru Strategic Program (SSP) survey over $\sim 1,140$ deg$^2$. In addition to three candidates that we already reported, we have identified two new MW satellite candidates in the constellation of Sextans at a heliocentric distance of $D_{\odot} \simeq 126$kpc, and Virgo at $D_{\odot} \simeq 151$kpc, named Sextans II and Virgo III, respectively. Their luminosities (Sext II:$M_V\simeq-3.9$mag; Vir III:$M_V\simeq-2.7$mag) and half-light radii (Sext II:$r_h\simeq154$ pc; Vir III:$r_h\simeq 44$ pc) place them in the region of size-luminosity space of ultra-faint dwarf galaxies (UFDs). Including four previously known satellites, there are a total of nine satellites in the HSC-SSP footprint. This discovery rate of UFDs is much higher than that predicted from the recent models for the expected population of MW satellites in the framework of cold dark matter models, thereby suggesting that we encounter a too many satellites problem. Possible solutions to settle this tension are also discussed., Comment: 20 pages, 17 figures, 4 tables, accepted for publication in PASJ

Large language models (LLMs) are accelerating the development of language-guided robot planners. Meanwhile, symbolic planners offer the advantage of interpretability. This paper proposes a new task that bridges these two trends, namely, multimodal planning problem specification. The aim is to generate a problem description (PD), a machine-readable file used by the planners to find a plan. By generating PDs from language instruction and scene observation, we can drive symbolic planners in a language-guided framework. We propose a Vision-Language Interpreter (ViLaIn), a new framework that generates PDs using state-of-the-art LLM and vision-language models. ViLaIn can refine generated PDs via error message feedback from the symbolic planner. Our aim is to answer the question: How accurately can ViLaIn and the symbolic planner generate valid robot plans? To evaluate ViLaIn, we introduce a novel dataset called the problem description generation (ProDG) dataset. The framework is evaluated with four new evaluation metrics. Experimental results show that ViLaIn can generate syntactically correct problems with more than 99\% accuracy and valid plans with more than 58\% accuracy. Our code and dataset are available at https://github.com/omron-sinicx/ViLaIn., Comment: ICRA 2024

The detection of gravitational waves from a binary neutron star merger by Advanced LIGO and Advanced Virgo (GW170817), along with the discovery of the electromagnetic counterparts of this gravitational wave event, ushered in a new era of multimessenger astronomy, providing the first direct evidence that BNS mergers are progenitors of short gamma-ray bursts (GRBs). Such events may also produce very-high-energy (VHE, > 100GeV) photons which have yet to be detected in coincidence with a gravitational wave signal. The Cherenkov Telescope Array (CTA) is a next-generation VHE observatory which aims to be indispensable in this search, with an unparalleled sensitivity and ability to slew anywhere on the sky within a few tens of seconds. New observing modes and follow-up strategies are being developed for CTA to rapidly cover localization areas of gravitational wave events that are typically larger than the CTA field of view. This work will evaluate and provide estimations on the expected number of of gravitational wave events that will be observable with CTA, considering both on- and off-axis emission. In addition, we will present and discuss the prospects of potential follow-up strategies with CTA., Comment: Presented at the 38th International Cosmic Ray Conference (ICRC 2023), 2023 (arXiv:2309.08219)

Paul traps are devices that confine particles using an alternating electric field and have been used in undergraduate experimental classes at universities. Owing to the requirement of a high voltage ($> 10^3$ V), Paul traps are not used in middle and high schools. Therefore, we developed an all-in-one-type Paul trap , including a high-voltage transformer. The Paul trap can be equipped with two different types of electrode attachments, ring-type and linear-type, and the trap image can be observed using a built-in web camera. For example, the charge-to-mass ratio of particles was measured with different types of attachments, and reasonable values were obtained. These types of trap devices are currently used at several educational facilities in Japan., Comment: 7 pages, 14 figures

The theoretically observable limit of electron density distribution by single-crystal X-ray diffraction is discussed. When F_{orb} and {\delta}F are defined as, respectively, the partial structure factor for an orbital and the deviation of the observed F from the true F, the accuracy of electron density attributable to F_{orb} is chiefly determined by the number of reflections satisfying the condition F_{orb}/F > {\delta}F/F. Since F_{orb}/F, which is generally small for crystals with large F(0,0,0), is constant under a given set of experimental conditions, {\delta}F/F must be reduced to increase the number of reflections satisfying F_{orb}/F > {\delta}F/F. The present paper demonstrates how to reduce {\delta}F mathematically and experimentally, and the following topics are covered: the Poisson statistics, accumulation of errors in the data collection and reduction procedure, multiple diffraction, conversion error from F^2 to F in refinement programs, which is unavoidable when the input quantities have different dimension from F, weighting of reflections, and tips. For demonstration, observation of the electron density of the Ti-3d^1 orbital in YTiO_3 by synchrotron single-crystal X-ray diffraction is presented., Comment: 68 pages, 20 figures

With the increase of the number of elderly people living alone around the world, there is a growing demand for sensor-based detection of anomalous behaviors. Although smart homes with ambient sensors could be useful for detecting such anomalies, there is a problem of lack of sufficient real data for developing detection algorithms. For coping with this problem, several sensor data simulators have been proposed, but they have not been able to model appropriately the long-term transitions and correlations between anomalies that exist in reality. In this paper, therefore, we propose a novel sensor data simulator that can model these factors in generation of sensor data. Anomalies considered in this study were classified into three types of \textit{state anomalies}, \textit{activity anomalies}, and \textit{moving anomalies}. The simulator produces 10 years data in 100 min. including six anomalies, two for each type. Numerical evaluations show that this simulator is superior to the past simulators in the sense that it simulates well day-to-day variations of real data., Comment: 26 pages, 10 figures

To assess the impact of potential future climate pledges after the first Global Stocktake, we propose a simple, transparent framework for developing emission and temperature scenarios by country. We show that current pledges with unconditional targets lead to global warming of 1.96 (1.39-2.6) degree C by 2100. Further warming could be limited through i) commitment to mid-century net-zero targets for all countries and earlier net-zero targets for developed countries, ii) enhancement of the Global Methane Pledge, and iii) ambitious implementation of the Glasgow Leaders Declaration on Forests and Land Use. Our analysis further shows that overshooting 1.5 degree C is unavoidable, even with supplementary climate engineering strategies, suggesting the need for strategies to limit further overshoot and ultimately reduce the warming towards 1.5 degree C.

A typical assumption in state-of-the-art self-localization models is that an annotated training dataset is available for the target workspace. However, this is not necessarily true when a robot travels around the general open world. This work introduces a novel training scheme for open-world distributed robot systems. In our scheme, a robot (``student") can ask the other robots it meets at unfamiliar places (``teachers") for guidance. Specifically, a pseudo-training dataset is reconstructed from the teacher model and then used for continual learning of the student model under domain, class, and vocabulary incremental setup. Unlike typical knowledge transfer schemes, our scheme introduces only minimal assumptions on the teacher model, so that it can handle various types of open-set teachers, including those uncooperative, untrainable (e.g., image retrieval engines), or black-box teachers (i.e., data privacy). In this paper, we investigate a ranking function as an instance of such generic models, using a challenging data-free recursive distillation scenario, where a student once trained can recursively join the next-generation open teacher set., Comment: 5 pages, 4 figures, technical report

Regulatory authorities guide the use of permutation tests or randomization tests so as not to increase the type-I error rate when applying covariate-adaptive randomization in randomized clinical trials. For non-inferiority and equivalence trials, this paper derives adjusted confidence intervals using permutation and randomization methods, thus controlling the type-I error to be much closer to the pre-specified nominal significance level. We consider three variable types for the outcome of interest, namely normal, binary, and time-to-event variables for the adjusted confidence intervals. For normal variables, we show that the type-I error for the adjusted confidence interval holds the nominal significance level. However, we highlight a unique theoretical challenge for non-inferiority and equivalence trials: binary and time-to-event variables may not hold the nominal significance level when the model parameters are estimated by models that diverge from the data-generating model under the null hypothesis. To clarify these features, we present simulation results and evaluate the performance of the adjusted confidence intervals.

The knot quandle is a complete invariant for oriented classical knots in the $3$-sphere up to orientation. Eisermann computed the second quandle homology group of the knot quandle and showed that it characterizes the unknot. In this paper, we compute the second quandle homology group of the knot $n$-quandle completely, where the knot $n$-quandle is a certain quotient of the knot quandle for each integer $n$ greater than one. Although the knot $n$-quandle is weaker than the knot quandle, the second quandle homology group of the former is found to have more information than that of the latter. As one of the consequences, it follows that the second quandle homology group of the knot $3$-quandle characterizes the unknot, the trefoil and the cinquefoil., Comment: 20 pages

With an X-ray stacking analysis of ~ 12, 000 Lyman-break galaxies (LBGs) using the Chandra Legacy Survey image, we investigate average supermassive black hole (SMBH) accretion properties of star-forming galaxies (SFGs) at 4 <~ z <~ 7. Although no X-ray signal is detected in any stacked image, we obtain strong 3 sigma upper limits for the average black hole accretion rate (BHAR) as a function of star formation rate (SFR). At z ~ 4 (5) where the stacked image is deeper, the 3 sigma BHAR upper limits per SFR are ~ 1.5 (1.0) dex lower than the local black hole-to-stellar mass ratio, indicating that the SMBHs of SFGs in the inactive (BHAR <~1M_sun yr^{-1}) phase are growing much more slowly than expected from simultaneous evolution. We obtain a similar result for BHAR per dark halo accretion rate. QSOs from the literature are found to have ~ 1 dex higher SFRs and >~ 2 dex higher BHARs than LBGs with the same dark halo mass. We also make a similar comparison for dusty starburst galaxies and quiescent galaxies from the literature. A duty-cycle corrected analysis shows that for a given dark halo, the SMBH mass increase in the QSO phase dominates over that in the much longer inactive phase. Finally, a comparison with the TNG300, TNG100, SIMBA100, and EAGLE100 simulations finds that they overshoot our BHAR upper limits by <~ 1.5 dex, possibly implying that simulated SMBHs are too massive., Comment: 16 pages, 11 figures, 3 tables. Accepted for publication in MNRAS

We search for the $e^+e^-\to\eta_b(1S)\omega$ and $e^+e^-\to\chi_{b0}(1P)\omega$ processes at a center-of-mass energy of 10.745 GeV, which is close to the peak of the $\Upsilon(10753)$ state. We use data collected by the Belle II experiment during a special run, corresponding to an integrated luminosity of $9.8\,\mathrm{fb}^{-1}$. We reconstruct $\omega\to\pi^+\pi^-\pi^0$ decays and use the $\omega$ meson's recoil mass to search for the signals. We do not find evidence for either process, and set upper limits on the corresponding Born-level cross sections of 2.5 pb and 7.8 pb, respectively, at the 90% confidence level. The $\chi_{b0}(1P)\omega$ limit is the result of a combination of this analysis and a previous search using full reconstruction., Comment: 8 pages, 3 figures, submitted to PRD

Sb-substituted CeOBiS2 single crystals with 0.2-1.0 mm size have been successfully grown using CsCl/KCl flux. Sb substitution strongly suppressed the superconductivity in CeOBiS2, and the substitution of more than approximately 4 atomic percent in Bi-site disappeared the superconductivity at a measurement range of above 0.36 K. Furthermore, the drastic change of c-axis lattice parameter was observed in a neighborhood to Sb substitution amount, which maybe suggests the structural transition in CeOBiS2. The Ce valence state of the obtained single crystals was evaluated. However, the established correlations between Ce valence state and superconductivity could not be revealed.