arXiv subjects

Dongha Kim

Publications and source records attributed to Dongha Kim.

At least 37 records · Page 2Linked to original sources

arXiv2025

DSAC: Low-Cost RowHammer Mitigation Using In-DRAM Stochastic and Approximate Counting Algorithm

This paper provides the fundamental mechanisms of two types of row activation-induced bit flips and proposes in-DRAM protection techniques. RowBleed occurs when a victim row experiences charge leakage due to transistor's threshold voltage lowering induced by long activation of a neighboring aggressor row. Therefore, this paper proposes Time-Weighted Counting for RowBleed mitigation, which assigns greater counter weights to rows that are activated for longer durations. On the other hand, RowHammer occurs when a victim row experiences electron injection due to frequent activation of a neighboring aggressor row. Similarly, Extended RowHammer, the phenomenon where victim rows are two rows beyond aggressor rows, is also caused by electron injection due to frequent activation of a neighboring aggressor row. Consequently, accurate detection of aggressor rows is crucial. Therefore, this paper proposes RowHammer mitigation algorithm named DSAC (in-DRAM Stochastic and Approximate Counting algorithm), which utilizes a replacement probability that adjusts based on the count of the old row. This paper introduces a RowHammer protection index called Maximum Disturbance, which measures the maximum accumulated number of row activations within an observation period. The experimental results demonstrate that DSAC can achieve 133x lower Maximum Disturbance than the state-of-the-art counter-based algorithm.

cs.CR↗

arXiv2025

Personalizing Large Language Models using Retrieval Augmented Generation and Knowledge Graph

The advent of large language models (LLMs) has allowed numerous applications, including the generation of queried responses, to be leveraged in chatbots and other conversational assistants. Being trained on a plethora of data, LLMs often undergo high levels of over-fitting, resulting in the generation of extra and incorrect data, thus causing hallucinations in output generation. One of the root causes of such problems is the lack of timely, factual, and personalized information fed to the LLM. In this paper, we propose an approach to address these problems by introducing retrieval augmented generation (RAG) using knowledge graphs (KGs) to assist the LLM in personalized response generation tailored to the users. KGs have the advantage of storing continuously updated factual information in a structured way. While our KGs can be used for a variety of frequently updated personal data, such as calendar, contact, and location data, we focus on calendar data in this paper. Our experimental results show that our approach works significantly better in understanding personal information and generating accurate responses compared to the baseline LLMs using personal data as text inputs, with a moderate reduction in response time.

cs.CL↗

arXiv2025

ALTBI: Constructing Improved Outlier Detection Models via Optimization of Inlier-Memorization Effect

Outlier detection (OD) is the task of identifying unusual observations (or outliers) from a given or upcoming data by learning unique patterns of normal observations (or inliers). Recently, a study introduced a powerful unsupervised OD (UOD) solver based on a new observation of deep generative models, called inlier-memorization (IM) effect, which suggests that generative models memorize inliers before outliers in early learning stages. In this study, we aim to develop a theoretically principled method to address UOD tasks by maximally utilizing the IM effect. We begin by observing that the IM effect is observed more clearly when the given training data contain fewer outliers. This finding indicates a potential for enhancing the IM effect in UOD regimes if we can effectively exclude outliers from mini-batches when designing the loss function. To this end, we introduce two main techniques: 1) increasing the mini-batch size as the model training proceeds and 2) using an adaptive threshold to calculate the truncated loss function. We theoretically show that these two techniques effectively filter out outliers from the truncated loss function, allowing us to utilize the IM effect to the fullest. Coupled with an additional ensemble strategy, we propose our method and term it Adaptive Loss Truncation with Batch Increment (ALTBI). We provide extensive experimental results to demonstrate that ALTBI achieves state-of-the-art performance in identifying outliers compared to other recent methods, even with significantly lower computation costs. Additionally, we show that our method yields robust performances when combined with privacy-preserving algorithms.

stat.ML↗

arXiv2025

Dynamic realization of emergent high-dimensional optical vortices

The dimensionality of vortical structures has recently been extended beyond two dimensions, providing higher-order topological characteristics and robustness for high-capacity information processing and turbulence control. The generation of high-dimensional vortical structures has mostly been demonstrated in classical systems through the complex interference of fluidic, acoustic, or electromagnetic waves. However, natural materials rarely support three- or higher-dimensional vortical structures and their physical interactions. Here, we present a high-dimensional gradient thickness optical cavity (GTOC) in which the optical coupling of planar metal-dielectric multilayers implements topological interactions across multiple dimensions. Topological interactions in high-dimensional GTOC construct non-trivial topological phases, which induce high-dimensional vortical structures in generalized parameter space in three, four dimensions, and beyond. These emergent high-dimensional vortical structures are observed under electro-optic tomography as optical vortex dynamics in two-dimensional real-space, employing the optical thicknesses of the dielectric layers as synthetic dimensions. We experimentally demonstrate emergent vortical structures, optical vortex lines and vortex rings, in a three-dimensional generalized parameter space and their topological transitions. Furthermore, we explore four-dimensional vortical structures, termed optical vortex sheets, which provide the programmability of real-space optical vortex dynamics. Our findings hold significant promise for emulating high-dimensional physics and developing active topological photonic devices.

physics.optics↗

arXiv2024

A Case Study of API Design for Interoperability and Security of the Internet of Things

Heterogeneous distributed systems, including the Internet of Things (IoT) or distributed cyber-physical systems (CPS), often suffer a lack of interoperability and security, which hinders the wider deployment of such systems. Specifically, the different levels of security requirements and the heterogeneity in terms of communication models, for instance, point-to-point vs. publish-subscribe, are the example challenges of IoT and distributed CPS consisting of heterogeneous devices and applications. In this paper, we propose a working application programming interface (API) and runtime to enhance interoperability and security while addressing the challenges that stem from the heterogeneity in the IoT and distributed CPS. In our case study, we design and implement our application programming interface (API) design approach using open-source software, and with our working implementation, we evaluate the effectiveness of our proposed approach. Our experimental results suggest that our approach can achieve both interoperability and security in the IoT and distributed CPS with a reasonably small overhead and better-managed software.

cs.DC↗

arXiv2024

Beyond Ontology in Dialogue State Tracking for Goal-Oriented Chatbot

Goal-oriented chatbots are essential for automating user tasks, such as booking flights or making restaurant reservations. A key component of these systems is Dialogue State Tracking (DST), which interprets user intent and maintains the dialogue state. However, existing DST methods often rely on fixed ontologies and manually compiled slot values, limiting their adaptability to open-domain dialogues. We propose a novel approach that leverages instruction tuning and advanced prompt strategies to enhance DST performance, without relying on any predefined ontologies. Our method enables Large Language Model (LLM) to infer dialogue states through carefully designed prompts and includes an anti-hallucination mechanism to ensure accurate tracking in diverse conversation contexts. Additionally, we employ a Variational Graph Auto-Encoder (VGAE) to model and predict subsequent user intent. Our approach achieved state-of-the-art with a JGA of 42.57% outperforming existing ontology-less DST models, and performed well in open-domain real-world conversations. This work presents a significant advancement in creating more adaptive and accurate goal-oriented chatbots.

cs.CL↗

arXiv2024

Illustrious: an Open Advanced Illustration Model

In this work, we share the insights for achieving state-of-the-art quality in our text-to-image anime image generative model, called Illustrious. To achieve high resolution, dynamic color range images, and high restoration ability, we focus on three critical approaches for model improvement. First, we delve into the significance of the batch size and dropout control, which enables faster learning of controllable token based concept activations. Second, we increase the training resolution of images, affecting the accurate depiction of character anatomy in much higher resolution, extending its generation capability over 20MP with proper methods. Finally, we propose the refined multi-level captions, covering all tags and various natural language captions as a critical factor for model development. Through extensive analysis and experiments, Illustrious demonstrates state-of-the-art performance in terms of animation style, outperforming widely-used models in illustration domains, propelling easier customization and personalization with nature of open source. We plan to publicly release updated Illustrious model series sequentially as well as sustainable plans for improvements.

cs.CV↗

arXiv2024

Shaping space-time wavepackets beyond the paraxial limit using a dispersion magnifier

Space-time wavepackets (STWPs) have received significant attention since they can propagate in free space at arbitrary group velocity without dispersion and diffraction. However, at present, the generation of STWPs has been limited to the paraxial regime. Here we show that conventional optical elements can be used to extend STWPs beyond the paraxial regime. A dispersion magnifier, consisting of two lenses and a beam expander, applies spatiotemporal shaping to paraxial STWPs to create nonparaxial STWPs. The control of the magnification ratio results in versatile engineering capabilities on group velocity, beam diameter, and propagation distance. As an example, we numerically demonstrated long-distance propagation or slow group velocity of the output wavepacket with subwavelength cross-sections.

physics.optics↗

arXiv2024

META-ANOVA: Screening interactions for interpretable machine learning

There are two things to be considered when we evaluate predictive models. One is prediction accuracy,and the other is interpretability. Over the recent decades, many prediction models of high performance, such as ensemble-based models and deep neural networks, have been developed. However, these models are often too complex, making it difficult to intuitively interpret their predictions. This complexity in interpretation limits their use in many real-world fields that require accountability, such as medicine, finance, and college admissions. In this study, we develop a novel method called Meta-ANOVA to provide an interpretable model for any given prediction model. The basic idea of Meta-ANOVA is to transform a given black-box prediction model to the functional ANOVA model. A novel technical contribution of Meta-ANOVA is a procedure of screening out unnecessary interaction before transforming a given black-box model to the functional ANOVA model. This screening procedure allows the inclusion of higher order interactions in the transformed functional ANOVA model without computational difficulties. We prove that the screening procedure is asymptotically consistent. Through various experiments with synthetic and real-world datasets, we empirically demonstrate the superiority of Meta-ANOVA

stat.ML↗

arXiv2024

ODIM: Outlier Detection via Likelihood of Under-Fitted Generative Models

The unsupervised outlier detection (UOD) problem refers to a task to identify inliers given training data which contain outliers as well as inliers, without any labeled information about inliers and outliers. It has been widely recognized that using fully-trained likelihood-based deep generative models (DGMs) often results in poor performance in distinguishing inliers from outliers. In this study, we claim that the likelihood itself could serve as powerful evidence for identifying inliers in UOD tasks, provided that DGMs are carefully under-fitted. Our approach begins with a novel observation called the inlier-memorization (IM) effect-when training a deep generative model with data including outliers, the model initially memorizes inliers before outliers. Based on this finding, we develop a new method called the outlier detection via the IM effect (ODIM). Remarkably, the ODIM requires only a few updates, making it computationally efficient-at least tens of times faster than other deep-learning-based algorithms. Also, the ODIM filters out outliers excellently, regardless of the data type, including tabular, image, and text data. To validate the superiority and efficiency of our method, we provide extensive empirical analyses on close to 60 datasets.

stat.ML↗

arXiv2024

Optimizing Quantum Convolutional Neural Network Architectures for Arbitrary Data Dimension

Quantum convolutional neural networks (QCNNs) represent a promising approach in quantum machine learning, paving new directions for both quantum and classical data analysis. This approach is particularly attractive due to the absence of the barren plateau problem, a fundamental challenge in training quantum neural networks (QNNs), and its feasibility. However, a limitation arises when applying QCNNs to classical data. The network architecture is most natural when the number of input qubits is a power of two, as this number is reduced by a factor of two in each pooling layer. The number of input qubits determines the dimensions (i.e. the number of features) of the input data that can be processed, restricting the applicability of QCNN algorithms to real-world data. To address this issue, we propose a QCNN architecture capable of handling arbitrary input data dimensions while optimizing the allocation of quantum resources such as ancillary qubits and quantum gates. This optimization is not only important for minimizing computational resources, but also essential in noisy intermediate-scale quantum (NISQ) computing, as the size of the quantum circuits that can be executed reliably is limited. Through numerical simulations, we benchmarked the classification performance of various QCNN architectures when handling arbitrary input data dimensions on the MNIST and Breast Cancer datasets. The results validate that the proposed QCNN architecture achieves excellent classification performance while utilizing a minimal resource overhead, providing an optimal solution when reliable quantum computation is constrained by noise and imperfections.

quant-ph↗

arXiv2024

Scaling behavior of the degree of circular polarization of surface plasmon polariton

Surface plasmon polaritons (SPPs) carry transverse optical spin within the evanescent field, which has enabled the demonstration of various chiral light-matter interactions in classical and quantum systems. To achieve high spin selectivity in the interactions, the elliptical polarization of the evanescent field should be made circular, but the engineering principle of the degree of circular polarization (DOCP) of SPPs has been lacking. In this study, we theoretically and numerically investigate the scaling behavior of the DOCP of the SPP field with respect to the modal effective refractive index (neff). The DOCP of the SPP field exhibits power-law scalability to the effective refractive index in the 1D layered system, regardless of the material, structural geometry, and excitation wavelength. The power-law scalability is also confirmed in 2D waveguide structures for in-plane and out-of-plane SPP fields, but the scaling exponents vary depending on the distance from the waveguide boundaries by the reduced symmetry of the given system. Due to Lorentz reciprocity, the power-law scalability can be extended to the coupling directionality of chiral emitters towards the plasmonic waveguide. To this end, we propose a chiral photonic platform for enhanced light-valley interaction, which utilizes simultaneous enhancement of the DOCP and coupling directionality. An incident SPP can excite a chiral emitter with high spin selectivity that unidirectionally couples the emitted light into the plasmonic waveguide depending on the valley polarization of excitons in 2D material. Our work provides a ground rule for designing chiral nanophotonic systems and paves the way for the exploration of scale-free phenomena of electromagnetic waves.

physics.optics↗

arXiv2023

A likelihood approach to nonparametric estimation of a singular distribution using deep generative models

We investigate statistical properties of a likelihood approach to nonparametric estimation of a singular distribution using deep generative models. More specifically, a deep generative model is used to model high-dimensional data that are assumed to concentrate around some low-dimensional structure. Estimating the distribution supported on this low-dimensional structure, such as a low-dimensional manifold, is challenging due to its singularity with respect to the Lebesgue measure in the ambient space. In the considered model, a usual likelihood approach can fail to estimate the target distribution consistently due to the singularity. We prove that a novel and effective solution exists by perturbing the data with an instance noise, which leads to consistent estimation of the underlying distribution with desirable convergence rates. We also characterize the class of distributions that can be efficiently estimated via deep generative models. This class is sufficiently general to contain various structured distributions such as product distributions, classically smooth distributions and distributions supported on a low-dimensional manifold. Our analysis provides some insights on how deep generative models can avoid the curse of dimensionality for nonparametric distribution estimation. We conduct a thorough simulation study and real data analysis to empirically demonstrate that the proposed data perturbation technique improves the estimation performance significantly.

stat.ML↗

arXiv2023

Broadband generation of quasi bound-state-in-continuum modes using subwavelength truncated cone resonators

Allowing a high quality factor (Q-factor) to a sub-wavelength dielectric resonator, quasi-bound states in the continuum (Q-BIC) has gained much interest. However, the Q-BIC resonance condition is too sensitive to the geometry of the resonator and its practical broadband generation on a single-wafer platform has been limited. Here, we present that employing the base angle as a structural degree of freedom, the truncated nano-cone resonator supports the Q-BIC resonance with a high Q-factor of >150 over a wide wavelength range of >100 nm. We expect our approach will boost the utilization of the Q-BIC resonance for various applications requiring broadband spectral tuning.

physics.optics↗

arXiv2023

Learning fair representation with a parametric integral probability metric

As they have a vital effect on social decision-making, AI algorithms should be not only accurate but also fair. Among various algorithms for fairness AI, learning fair representation (LFR), whose goal is to find a fair representation with respect to sensitive variables such as gender and race, has received much attention. For LFR, the adversarial training scheme is popularly employed as is done in the generative adversarial network type algorithms. The choice of a discriminator, however, is done heuristically without justification. In this paper, we propose a new adversarial training scheme for LFR, where the integral probability metric (IPM) with a specific parametric family of discriminators is used. The most notable result of the proposed LFR algorithm is its theoretical guarantee about the fairness of the final prediction model, which has not been considered yet. That is, we derive theoretical relations between the fairness of representation and the fairness of the prediction model built on the top of the representation (i.e., using the representation as the input). Moreover, by numerical experiments, we show that our proposed LFR algorithm is computationally lighter and more stable, and the final prediction model is competitive or superior to other LFR algorithms using more complex discriminators.

stat.ML↗

arXiv2022

Customising radiative decay dynamics of two-dimensional excitons via position- and polarisation-dependent vacuum-field interference

Embodying bosonic and electrically interactive characteristics in two-dimensional space, excitons in transition-metal dichalcogenides (TMDCs) have garnered considerable attention. The realisation and application of strong-correlation effects, long-range transport, and valley-dependent optoelectronic properties require customising exciton decay dynamics. Strains, defects, and electrostatic doping effectively control the decay dynamics but significantly disturb the intrinsic properties of TMDCs, such as electron band structure and exciton binding energy. Meanwhile, vacuum-field manipulation provides an optical alternative for engineering radiative decay dynamics. Planar mirrors and cavities have been employed to manage the light-matter interactions of two-dimensional excitons. However, the conventional flat platforms cannot customise the radiative decay landscape in the horizontal TMDC plane or independently control vacuum field interference at different pumping and emission frequencies. Here, we present a meta-mirror resolving the issues with more optical freedom. For neutral excitons of the monolayer MoSe2, the meta-mirror manipulated the radiative decay rate by two orders of magnitude, depending on its geometry. Moreover, we experimentally identified the correlation between emission intensity and spectral linewidth. The anisotropic meta-mirror demonstrated polarisation-dependent radiative decay control. We expect that the meta-mirror platform will be promising to tailor the two-dimensional distributions of lifetime, density, and diffusion of TMDC excitons in advanced opto-excitonic applications.

cond-mat.mes-hall↗

arXiv2022

Spontaneous generation and active manipulation of real-space optical vortex

Optical vortices host the orbital nature of photons, which offers an extra degree of freedom in photonic applications. Unlike vortices in other physical entities, optical vortices require structural singularities, which restrict their abilities in terms of dynamic and interactive characteristics. In this study, we present the spontaneous generation and external magnetic field-induced manipulation of an optical vortex and antivortex. A gradient-thickness optical cavity (GTOC) consisting of an Al/SiO2/Ni/SiO2 multilayer structure realised the distinct transition between the trivial and non-trivial topological phases, depending on the magneto-optic effects of the Ni layer. In the non-trivial topological phase, the mathematical singularities generating the optical vortex and antivortex pair in the reflected light existed in the generalised parameter space of the thicknesses of the top and bottom SiO2 layers, which is bijective to the real space of the GTOC. Coupled with the magnetisation, the optical vortex and antivortex in the GTOC experienced an effective spin-orbit interaction and showed topology-dependent dynamics under external magnetic fields. We expect that field-induced engineering of optical vortices will pave the way for the study of topological photonic interactions and their applications.

physics.optics↗

arXiv2021

Plasmonic photonic crystal mirror for long-lived interlayer exciton generation

Interlayer excitons in van der Waals heterostructures of two-dimensional transition metal dichalcogenides have recently emerged as a fascinating platform for quantum many-body effects, long-range interactions, and opto-valleytronic applications. The practical implementation of such phenomena and applications requires further development of the long-lived character of interlayer excitons. Whereas material developments have successfully enhanced the nonradiative lifetime, the out-of-plane polarization nature of the interlayer excitons has made it challenging to improve the radiative lifetime with conventional photonic mirrors. Here, we propose and systematically analyze a plasmonic photonic crystal (PPhC) mirror that can increase the radiative lifetime of interlayer excitons by two orders of magnitude. Based on the vacuum field transition, the PPhC mirror supports spatially uniform radiative decay suppression over its territory, which is crucial for engineering the interlayer excitons not localized at a specific position. The PPhC mirror platform will offer new possibilities for realizing long-lived interlayer exciton-based nanodevices.

cond-mat.mes-hall↗