Explore the vast range of titles available.

Since the 1987 democratic reform, regional attitudes toward the two southern provinces have been regarded as the most significant determinant of South Korean voters' political preferences. In recent years, however, many scholars have speculated that regional animus is losing its ground as a basis for political judgment and will eventually be replaced by other politically relevant factors such as political ideology and issue preference. Others raise questions about the validity of this kind of \"revisionist\" argument, noting that election outcomes remain regionally divided. Adopting the implicit association test (IAT) as a measure of regional attitudes, we provide an empirical test to measure the effects of the emergence of new generations and geographical mobility in South Korean voters' regional attitudes. Our results show that younger Korean voters are less regionally biased toward either of the two southern provinces. The only exception was the younger generation in the Honam region: they held even higher levels of animus toward Yeongnam than their older counterparts. Geographical mobility also seems to decrease regional bias. Those who have relocated to other provinces showed less regional bias when compared with the natives of Yeongnam and Honam still residing in their respective regions. Likewise, our results show that the inter-generational transfer of regional animus is not overwhelming. For the descendants of Yeongnam or Honam natives residing in another province, their family's region of origin mattered little. In short, our findings suggest that changing generational composition and geographical mobility are likely to lessen the severity of regional animus and the political significance of regional attitudes will wane.

The impressive development of large language models (LLMs) is expanding into the realm of large multimodal models (LMMs), which incorporate multiple types of data beyond text. However, the nature of multimodal models leads to significant expenses in the creation of training data. Furthermore, constructing multilingual data for LMMs presents its own set of challenges due to language diversity and complexity. Therefore, in this study, we propose two cost-effective methods to solve this problem: (1) vocabulary expansion and pretraining of multilingual LLM for specific languages, and (2) automatic and elaborate construction of multimodal datasets using GPT4-V. Based on015 these methods, we constructed a 91K English-Korean-Chinese multilingual, multimodal training dataset. Additionally, we developed a bilingual multimodal model that exhibits excellent performance in both Korean and English, surpassing existing approaches.

We propose the VLR-Bench, a visual question answering (VQA) benchmark for evaluating vision language models (VLMs) based on retrieval augmented generation (RAG). Unlike existing evaluation datasets for external knowledge-based VQA, the proposed VLR-Bench includes five input passages. This allows testing of the ability to determine which passage is useful for answering a given query, a capability lacking in previous research. In this context, we constructed a dataset of 32,000 automatically generated instruction-following examples, which we denote as VLR-IF. This dataset is specifically designed to enhance the RAG capabilities of VLMs by enabling them to learn how to generate appropriate answers based on input passages. We evaluated the validity of the proposed benchmark and training data and verified its performance using the state-of-the-art Llama3-based VLM, the Llava-Llama-3 model. The proposed VLR-Bench and VLR-IF datasets are publicly available online.

Large language models (LLMs) use pretraining to predict the subsequent word; however, their expansion requires significant computing resources. Numerous big tech companies and research institutes have developed multilingual LLMs (MLLMs) to meet current demands, overlooking less-resourced languages (LRLs). This study proposed three strategies to enhance the performance of LRLs based on the publicly available MLLMs. First, the MLLM vocabularies of LRLs were expanded to enhance expressiveness. Second, bilingual data were used for pretraining to align the high- and less-resourced languages. Third, a high-quality small-scale instruction dataset was constructed and instruction-tuning was performed to augment the LRL. The experiments employed the Llama2 model and Korean was used as the LRL, which was quantitatively evaluated against other developed LLMs across eight tasks. Furthermore, a qualitative assessment was performed based on human evaluation and GPT4. Experimental results showed that our proposed Bllossom model exhibited superior performance in qualitative analyses compared to previously proposed Korean monolingual models.

State monitoring of the complex system needs a large number of sensors. Especially, studies in soft electronics aim to attain complete measurement of the body, mapping various stimulations like temperature, electrophysiological signals, and mechanical strains. However, conventional approach requires many sensor networks that cover the entire curvilinear surfaces of the target area. We introduce a new measuring system, a novel electronic skin integrated with a deep neural network that captures dynamic motions from a distance without creating a sensor network. The device detects minute deformations from the unique laser-induced crack structures. A single skin sensor decodes the complex motion of five finger motions in real-time, and the rapid situation learning (RSL) ensures stable operation regardless of its position on the wrist. The sensor is also capable of extracting gait motions from pelvis. This technology is expected to provide a turning point in health-monitoring, motion tracking, and soft robotics. Real-time monitoring human motions normally demands connecting a large number of sensors in a complicated network. To make it simpler, Kim et al. decode the motion of fingers using a flexible sensor attached on wrist that measures skin deformation with the help of a deep-learning architecture.

Optoelectronic synapses can perceive and memorize visual information, making them appealing for future bionic eyes or vision automation. Organic field-effect transistors are a promising platform for optoelectronic synaptic devices thanks to their flexibility and biocompatibility. However, charge screening effects occurring at channel–dielectric interfaces hinder the implementation of programmable multilevel memories. Here, we report photonic organic synapses based on photon-modulated electrochemical doping in electrochemical transistors, where light can manipulate ion insertion into the photoactive layer composed of donor–acceptor heterojunction interfaces. This enables high-density multilevel conductance modulation at low operating voltages (<1 V) and the imitation of ion flux-driven synaptic activity of living systems. The devices can recognize different optical signals and mimic the learning processes of the human brain. By exploiting the integrated functions of perception, processing and memorization of visual information, a single-layer synapse array acts as an artificial retina enabling facial recognition without the use of a complex artificial neural network.An optoelectronic synapse is realized by incorporating a photoactive layer in an organic electrochemical transistor. Writing and erasing multiple conductance states allow optical signals to be recognized and the learning process of the human brain to be mimicked.

Pressure-sensitive touch panels can measure pressure and location (3D) information simultaneously and provide an intuitive and natural method for expressing one’s intention with a higher level of controllability and interactivity. However, they have been generally realized by a simple combination of pressure and location sensor or a stylus-based interface, which limit their implementation in a wide spectrum of applications. Here, we report a first demonstration (to our knowledge) of a transparent and flexible 3D touch which can sense the 3D information in a single device with the assistance of functionally designed self-generated multiscale structures. The single 3D touch system is demonstrated to draw a complex three-dimensional structure by utilizing the pressure as a third coordinate. Furthermore, rigorous theoretical analysis is carried out to achieve the target pressure performances with successful 3D data acquisition in wireless and wearable conditions, which in turn, paves the way for future wearable devices. Touch technology holds potential for the development of smartphones and touchscreens, yet the conventional devices are usually built on separate pressure and location sensing units. Kim et al. show a flexible and transparent touch sensor capable of mapping the position and pressure at the same time.

Non-alcoholic fatty liver disease (NAFLD) is currently one of the most common types of chronic liver injury. Elevated serum uric acid is a strong predictor of the development of fatty liver as well as metabolic syndrome. Here we demonstrate that uric acid induces triglyceride accumulation by SREBP-1c activation via induction of endoplasmic reticulum (ER) stress in hepatocytes. Uric acid-induced ER stress resulted in an increase of glucose-regulated protein (GRP78/94), splicing of the X-box-binding protein-1 (XBP-1), the phosphorylation of protein kinase RNA-like ER kinase (PERK), and eukaryotic translation initiation factor-2 α (eIF-2 α ) in cultured hepatocytes. Uric acid promoted hepatic lipogenesis through overexpression of the lipogenic enzyme, acetyl-CoA carboxylase 1 (ACC1), fatty acid synthase (FAS), and stearoyl-CoA desaturase 1 (SCD1) via activation of SREBP-1c, which was blocked by probenecid, an organic anion transport blocker in HepG2 cells and primary hepatocytes. A blocker of ER stress, tauroursodeoxycholic acid (TUDCA), and an inhibitor of SREBP-1c, metformin, blocked hepatic fat accumulation, suggesting that uric acid promoted fat synthesis in hepatocytes via ER stress-induced activation of SREBP-1c. Uric acid-induced activation of NADPH oxidase preceded ER stress, which further induced mitochondrial ROS production in hepatocytes. These studies provide new insights into the mechanisms by which uric acid stimulates fat accumulation in the liver.

Electrochromic (EC) windows face trade-offs in cost, switching speed, color neutrality, and durability. We present solution-processed n-doped poly(benzodifurandione) (n-PBDF) as a robust organic conductor for black EC windows overcoming these limitations. Utilizing an engineered solvent ink and ultrasonic spray coating, we achieve uniform large-area n-PBDF deposition under ambient conditions without additives. n-PBDF EC electrodes show unprecedented weathering durability (maintaining performance under simultaneous exposure to light (including UV), heat, and humidity for >768 h), addressing a key barrier for organic EC materials. The electrodes exhibit deep black coloration with color neutrality, rapid switching (<2 s), and remarkable cycling stability (>20,000 cycles). Large-area EC devices demonstrate uniform switching performance, confirming scalable fabrication. Building energy simulations of the EC window reveal significant HVAC savings potential across diverse transitional climates. This work establishes n-PBDF as a scalable, high-performance alternative to conventional inorganic EC systems, advancing the viability of solution-processable smart windows for sustainable architecture. Electrochromic materials are widely explored in energy-saving smart windows, yet combining fast switching, neutral black coloration, and robust long-term durability remains challenging. Here the authors report a solution processed n-doped poly(benzodifurandione) affording an electrochromic black window that overcomes these limitations.

Mid-infrared wavelengths are called the molecular fingerprint region because it contains the fundamental vibrational modes inherent to the substances of interest. Since the mid-infrared spectrum can provide non-destructive identification and quantitative analysis of unknown substances, miniaturized mid-infrared spectrometers for on-site diagnosis have attained great concern. Filter-array based on-chip spectrometer has been regarded as a promising alternative. In this study, we explore a way of applying a pillar-type plasmonic nanodiscs array, which is advantageous not only for excellent tunability of resonance wavelength but also for 2-dimensional integration through a single layer process, to the multispectral filter array for the on-chip spectrometer. We theoretically and experimentally investigated the optical properties of multi-periodic triangular lattices of metal nanodiscs array that act as stopband filters in the mid-infrared region. Soft-mold reverse nanoimprint lithography with a subsequent lift-off process was employed to fabricate the multispectral filter array and its filter function was successfully extracted using a Fourier transform infrared microscope. With the measured filter function, we tested the feasibility of target spectrum reconstruction using a Tikhonov regularization method for an ill-posed linear problem and evaluated its applicability to the infrared spectroscopic sensor that monitors an oil condition. These results not only verify that the multispectral filter array composed of stopband filters based on the metal nanodiscs array when combined with the spectrum reconstruction technique, has great potential for use to a miniaturized mid-infrared on-chip spectrometer, but also provide effective guidance for the filter design.