arXiv subjects

Wonjun Choi

Publications and source records attributed to Wonjun Choi.

18 recordsLinked to original sources

arXiv2026

PGMem: Tightly Coupled Persona-Memory Graph for Lifelong Personalized Agents

Long-term personalized dialogue agents must track user preferences as their personas evolve. Existing memory systems organize past events well, but store personas as flat profiles detached from the events that justify them. This loose coupling leads to the memory-persona validity gap and the persona-aware retrieval gap. We propose PGMem, a heterogeneous persona-memory graph that connects event and persona nodes through typed provenance and evidence edges, keeping each persona signal traceable to the events that support or revise it. At retrieval time, PGMem expands from query-relevant seeds and ranks signals by evidential validity. Across three benchmarks with small language model backbones, PGMem consistently outperforms summary-based, persona-aware, graph-structured, and agentic memory baselines, and improves performance as the context grows. The source code of PGMem is available at https://github.com/wonjunchoi23/pgmem/

cs.CL↗

arXiv2026

Can Structural Cues Save LLMs? Evaluating Language Models in Massive Document Streams

Evaluating language models in streaming environments is critical, yet underexplored. Existing benchmarks either focus on single complex events or provide curated inputs for each query, and do not evaluate models under the conflicts that arise when multiple concurrent events are mixed within the same document stream. We introduce StreamBench, a benchmark built from major news stories in 2016 and 2025, comprising 605 events and 15,354 documents across three tasks: Topic Clustering, Temporal Question Answering, and Summarization. To diagnose how models fail, we compare performance with and without structural cues, which organize key facts by event. We find that structural cues improve performance on clustering (up to +4.37%) and temporal QA (up to +9.63%), helping models locate relevant information and separate distinct events. While temporal reasoning remains an open challenge inherent to current LLMs, consistent gains across tasks show that structural cues are a promising direction for future work in massive document streams.

cs.CL↗

arXiv2024

Integrating Query-aware Segmentation and Cross-Attention for Robust VQA

This paper introduces a method for VizWiz-VQA using LVLM with trainable cross-attention and LoRA finetuning. We train the model with the following conditions: 1) Training with original images. 2) Training with enhanced images using CLIPSeg to highlight or contrast the original image. 3) Training with integrating the output features of Vision Transformer (ViT) and CLIPSeg features of the original images. Then, we ensemble the results based on Levenshtein distance to enhance the prediction of the final answer. In the experiments, we demonstrate and analyze the proposed method's effectiveness.

cs.CV↗

arXiv2024

Ptychographic lensless coherent endomicroscopy through a flexible fiber bundle

Conventional fiber-bundle-based endoscopes allow minimally invasive imaging through flexible multi-core fiber (MCF) bundles by placing a miniature lens at the distal tip and using each core as an imaging pixel. In recent years, lensless imaging through MCFs was made possible by correcting the core-to-core phase distortions pre-measured in a calibration procedure. However, temporally varying wavefront distortions, for instance, due to dynamic fiber bending, pose a challenge for such approaches. Here, we demonstrate a coherent lensless imaging technique based on intensity-only measurements insensitive to core-to-core phase distortions. We leverage a ptychographic reconstruction algorithm to retrieve the phase and amplitude profiles of reflective objects placed at a distance from the fiber tip, using as input a set of diffracted intensity patterns reflected from the object when the illumination is scanned over the MCF cores. Our approach thus utilizes an acquisition process equivalent to confocal microendoscopy, only replacing the single detector with a camera.

physics.optics↗

arXiv2023

BackTrack: Robust template update via Backward Tracking of candidate template

Variations of target appearance such as deformations, illumination variance, occlusion, etc., are the major challenges of visual object tracking that negatively impact the performance of a tracker. An effective method to tackle these challenges is template update, which updates the template to reflect the change of appearance in the target object during tracking. However, with template updates, inadequate quality of new templates or inappropriate timing of updates may induce a model drift problem, which severely degrades the tracking performance. Here, we propose BackTrack, a robust and reliable method to quantify the confidence of the candidate template by backward tracking it on the past frames. Based on the confidence score of candidates from BackTrack, we can update the template with a reliable candidate at the right time while rejecting unreliable candidates. BackTrack is a generic template update scheme and is applicable to any template-based trackers. Extensive experiments on various tracking benchmarks verify the effectiveness of BackTrack over existing template update algorithms, as it achieves SOTA performance on various tracking benchmarks.

cs.CV↗

arXiv2022

Wave propagation dynamics inside a complex scattering medium by the temporal control of backscattered waves

Shaping the wavefront of an incident wave to a complex scattering medium has demonstrated interesting possibilities, such as sub-diffraction wave focusing and enhancing light energy delivery. However, wavefront shaping has mainly been based on the control of transmitted waves that are inaccessible in most realistic applications. Here, we investigate the effect of maximizing the backscattered waves at a specific flight time on wave propagation dynamics and energy transport. We find both experimentally and numerically that the maximization at a short flight time focuses waves on the particles constituting the scattering medium, leading to the attenuation of the wave transport. On the contrary, maximization at a long flight time induces constructive wave interference inside the medium and thus enhances wave transport. We provide a theoretical model explaining this interesting transition behavior based on wave correlation. Our study provides a fundamental understanding of the effect of wave control on internal wave dynamics.

physics.optics↗

arXiv2021

A hybrid percolation transition at a finite transition point in scale-free networks

Percolation transition (PT) means the formation of a macroscopic-scale large cluster, which exhibits a continuous transition. However, when the growth of large clusters is globally suppressed, the type of PT is changed to a discontinuous transition for random networks. A question arises as to whether the type of PT is also changed for scale-free (SF) network, because the existence of hubs incites the formation of a giant cluster. Here, we apply a global suppression rule to the static model for SF networks, and investigate properties of the PT. We find that even for SF networks with the degree exponent $2 < \lambda <3$, a hybrid PT occurs at a finite transition point $t_c$, which we can control by the suppression strength. The order parameter jumps at $t_c^-$ and exhibits a critical behavior at $t_c^+$.

cond-mat.stat-mech↗

arXiv2020

Fourier holographic endoscopy for label-free imaging through a narrow and curved passage

Ultrathin lensless fibre endoscopes offer minimally invasive investigation, but they mostly operate as a rigid type due to the need for prior calibration of a fibre probe. Furthermore, most implementations work in fluorescence mode rather than label-free imaging mode, making them unsuitable for medicine and industry. Herein, we report a fully flexible ultrathin fibre endoscope taking 3D holographic images of unstained tissues with 0.87-{\mu}m spatial resolution. Using a bare fibre bundle as thin as 200-{\mu}m diameter, we design a lensless Fourier holographic imaging configuration to selectively detect weak reflections from biological tissues, a critical step for stain-free reflectance imaging. A unique algorithm is developed for calibration-free holographic image reconstruction, allowing us to image through a narrow and curved passage regardless of fibre bending. We demonstrate endoscopic reflectance imaging of unstained rat intestine tissues that are completely invisible to conventional endoscopes. The proposed endoscope will expedite more accurate and earlier diagnosis than before with minimal complications.

physics.optics↗

arXiv2019

Near-field imaging beyond the probe aperture limit

Near-field scanning optical microscopy has been an indispensable tool for designing, characterizing and understanding the functionalities of diverse nanoscale photonic devices. As the advances in fabrication technology have driven the devices smaller and smaller, the demand has grown steadily for improving its resolving power, which is determined mainly by the size of the probe attached to the scanner. The use of a smaller probe has been a straightforward approach to increase the resolving power, but it cannot be made arbitrarily small in practice due to the steep reduction of the collection efficiency. Here, we develop a method to enhance the resolving power of near-field imaging beyond the limit set by the physical size of the probe aperture. The main working principle is to unveil high-order near-field eigenmodes invisible with conventional near-field microscopy. The destructive interference of near-field waves is induced in these high-order eigenmodes by the locally varying phases, which can reveal subaperture-scale fine structural details. To extract these eigenmodes, we construct a self-interference near-field microscopy system and measure a fully phase-referenced far- to near-field transmission matrix (FNTM) composed of near-field amplitude and phase maps recorded for various angles of far-field illumination. By the singular value decomposition of the measured FNTM, we could extract the antisymmetric mode, quadrupole mode, and other higher-order modes hidden under the lowest-order symmetric mode. This enables us to resolve double and triple nano-slots whose gap size (50 nm) is three times smaller than the diameter of the probe aperture (150 nm). The subaperture near-field mode mapping by the FTNM can be potentially combined with various existing near-field imaging modalities and promote their ability to interrogate local near-field optical waves of nanoscale devices.

physics.optics↗

arXiv2017

Focusing of light energy inside a scattering medium by controlling the time-gated multiple light scattering

The efficient delivery of light energy is a prerequisite for non-invasive imaging and stimulating of target objects embedded deep within a scattering medium. However, injected waves experience random diffusion by multiple light scattering, and only a small fraction reaches the target object. Here we present a method to counteract wave diffusion and to focus multiplescattered waves to the deeply embedded target. To realize this, we experimentally inject light to the reflection eigenchannels of a specific flight time where most of the multiple-scattered waves have interacted with the target object and maximize the intensity of the returning multiple-scattered waves at the selected time. For targets that are too deep to be visible by optical imaging, we demonstrated a more than 10-fold enhancement in light energy delivery in comparison with ordinary wave diffusion cases. This work will lay a foundation for enhancing the working depth of imaging, sensing, and light stimulation.

physics.optics↗

arXiv2017

Absence of Anderson localization in certain random lattices

We report on the transition between an Anderson localized regime and a conductive regime in a 1D scattering system with correlated disorder. We show experimentally that when long-range correlations, in the form of a power-law spectral density with power larger than 2, are introduced the localization length becomes much bigger than the sample size and the transmission peaks typical of an Anderson localized system merge into a pass band. As other forms of long-range correlations are known to have the opposite effect, i.e. to enhance localization, our results show that care is needed when discussing the effects of correlations, as different kinds of long-range correlations can give rise to very different behavior.

physics.optics↗

arXiv2017

Two golden times in two-step contagion models

The two-step contagion model is a simple toy model for understanding pandemic outbreaks that occur in the real world. The model takes into account that a susceptible person either gets immediately infected or weakened when getting into contact with an infectious one. As the number of weakened people increases, they eventually can become infected in a short time period and a pandemic outbreak occurs. The time required to reach such a pandemic outbreak allows for intervention and is often called golden time. Understanding the size-dependence of the golden time is useful for controlling pandemic outbreak. Here we find that there exist two types of golden times in the two-step contagion model, which scale as $O(N^{1/3})$ and $O(N^{\zeta})$ with the system size $N$ on Erd\H{o}s-R\'enyi networks, where the measured $\zeta$ is slightly larger than $1/4$. They are distinguished by the initial number of infected nodes, $o(N)$ and $O(N)$, respectively. While the exponent $1/3$ of the $N$-dependence of the golden time is universal even in other models showing discontinuous transitions induced by cascading dynamics, the measured $\zeta$ exponents are all close to $1/4$ but show model-dependence. It remains open whether or not $\zeta$ reduces to $1/4$ in the asymptotically large-$N$ limit.

q-bio.PE↗

arXiv2017

Critical behavior of a two-step contagion model with multiple seeds

A two-step contagion model with a single seed serves as a cornerstone for understanding the critical behaviors and underlying mechanism of discontinuous percolation transitions induced by cascade dynamics. When the contagion spreads from a single seed, a cluster of infected and recovered nodes grows without any cluster merging process. However, when the contagion starts from multiple seeds of $O(N)$ where $N$ is the system size, a node weakened by a seed can be infected more easily when it is in contact with another node infected by a different pathogen seed. This contagion process can be viewed as a cluster merging process in a percolation model. Here, we show analytically and numerically that when the density of infectious seeds is relatively small but $O(1)$, the epidemic transition is hybrid, exhibiting both continuous and discontinuous behavior, whereas when it is sufficiently large and reaches a critical point, the transition becomes continuous. We determine the full set of critical exponents describing the hybrid and the continuous transitions. Their critical behaviors differ from those in the single-seed case.

cond-mat.stat-mech↗

arXiv2017

Universal mechanism for hybrid percolation transitions

Hybrid percolation transitions (HPTs) induced by cascading processes have been observed in diverse complex systems such as $k$-core percolation, breakdown on interdependent networks and cooperative epidemic spreading models. Much effort has been devoted to describe the properties of HPTs of individual systems. Yet the fundamental question about the possible universal mechanism underlying those HPTs has not been investigated at a microscopic level. Here, we find that the discontinuity in the order parameter in such HPTs results from two steps: a durable critical branching (CB) and an explosive, supercritical (SC) process. In a random network of $N$ nodes at the transition the CB process persists for $O(N^{1/3})$ time and the remaining nodes become vulnerable. Those vulnerable nodes are activated then in the short SC process. This crossover mechanism and scaling behavior are universal for different HPT systems.

physics.soc-ph↗

arXiv2016

Mixed-order phase transition in a two-step contagion model with single infectious seed

A hybrid phase transition (HPT) that exhibits properties of continuous and discontinuous phase transitions at the same transition point has been observed in diverse complex systems. Previous studies of the HPTs on complex networks mainly focused on whether the order parameter is continuous or discontinuous. However, more careful and fundamental questions on the critical behaviors of the HPT such as how the divergences of the susceptibility and of the correlation size are affected by the discontinuity of the order parameter have been addressed. Here, we consider a generalized epidemic model that is known to exhibit a discontinuous transition as a spinodal transition. Performing extensive numerical simulations and using finite-size scaling analysis, we examine diverging behaviors of the susceptibility and the correlation size. We find that when there is one infectious node and under a certain condition, the order parameter can exhibit a discontinuous jump but does not exhibit any critical behavior before or after the jump. This feature differs from what we observed in HPTs in the percolation pruning process. However, critical behavior appears in the form of a power-law behavior of the outbreak size distribution. The mean outbreak size, corresponding to the susceptibility, diverge following the conventional percolation behavior. Thus a mixed-order transition occurs. The hyperscaling relation does not hold.

cond-mat.stat-mech↗

arXiv2015

Hot Carrier Trapping Induced Negative Photoconductance in InAs Nanowires toward Novel Nonvolatile Memory

We report a novel negative photoconductivity (NPC) mechanism in n-type indium arsenide nanowires (NWs). Photoexcitation significantly suppresses the conductivity with a gain up to 10^5. The origin of NPC is attributed to the depletion of conduction channels by light assisted hot electron trapping, supported by gate voltage threshold shift and wavelength dependent photoconductance measurements. Scanning photocurrent microscopy excludes the possibility that NPC originates from the NW/metal contacts and reveals a competing positive photoconductivity. The conductivity recovery after illumination substantially slows down at low temperature, indicating a thermally activated detrapping mechanism. At 78 K, the spontaneous recovery of the conductance is completely quenched, resulting in a reversible memory device which can be switched by light and gate voltage pulses. The novel NPC based optoelectronics may find exciting applications in photodetection and nonvolatile memory with low power consumption.

cond-mat.mes-hall↗

arXiv2013

Exploring anti-reflection modes in disordered media

Sensing and manipulating targets hidden under scattering media are universal problems that take place in applications ranging from deep-tissue optical imaging to laser surgery. A major issue in these applications is the shallow light penetration caused by multiple scattering that reflects most of incident light. Although advances have been made to eliminate image distortion by a scattering medium, dealing with the light reflection has remained unchallenged. Here we present a method to minimize reflected intensity by finding and coupling light into the anti-reflection modes of a scattering medium. In doing so, we achieved more than a factor of 3 increase in light penetration. Our method of controlling reflected waves makes it readily applicable to in vivo applications in which detector sensors can only be positioned at the same side of illumination and will therefore lay the foundation of advancing the working depth of many existing optical imaging and treatment technologies.

physics.optics↗

arXiv2013

Enrichment of deeply penetrating waves in disordered media

Waves incident to a highly scattering medium are incapable of penetrating deep into the medium due to the diffusion process induced by multiple scattering. This poses a fundamental limitation to optically imaging, sensing, and manipulating targets embedded in opaque scattering layers such as biological tissues. One strategy for mitigating the shallow wave penetration is to exploit eigenmodes with anomalously high transmittance existing in any disordered medium. When waves are coupled to these eigenmodes, strong constructive wave interference enhances deeply penetrating waves. However, finding such eigenmodes has been a challenging task due to the complexity of disordered media. In this Letter, we present an iterative wavefront control method that selectively enriches the coupling of incident beam to high-transmission eigenmodes. Specifically, we refined the high-transmission eigenmodes from an arbitrary initial wave by either maximizing transmitted wave intensity or minimizing reflected wave intensity. Using the proposed method, we achieved more than a factor of 3 increases in light transmission through a scattering medium exhibiting hundreds of scattering events. Our approach is readily applicable to in vivo applications in which only the detection of reflected waves is available. Enhancing light penetration will lead to improving the working depth of optical imaging and treatment techniques.

physics.optics↗