Source author record

Seunghyun Lee

Seunghyun Lee appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Catalog footprint

What is connected

27works

21topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

Automated Visualization Code Synthesis via Multi-Path Reasoning and Feedback-Driven Optimization

Large Language Models (LLMs) have become a cornerstone for automated visualization code generation, enabling users to create charts through natural language instructions. Despite improvements from techniques like few-shot prompting and query expansion, existing methods often struggle when requests are underspecified in actionable details (e.g., data preprocessing assumptions, solver or library choices, etc.), frequently necessitating manual intervention. To overcome these limitations, we propose VisPath: a Multi-Path Reasoning and Feedback-Driven Optimization Framework for Visualization Code Generation. VisPath handles underspecified queries through structured, multi-stage processing. It begins by using Chain-of-Thought (CoT) prompting to reformulate the initial user input, generating multiple extended queries in parallel to surface alternative plausible concretizations of the request. These queries then generate candidate visualization scripts, which are executed to produce diverse images. By assessing the visual quality and correctness of each output, VisPath generates targeted feedback that is aggregated to synthesize an optimal final result. Extensive experiments on MatPlotBench and Qwen-Agent Code Interpreter Benchmark show that VisPath outperforms state-of-the-art methods, providing a more reliable framework for AI-driven visualization generation.

preprint2026arXiv

ExploitBench: A Capability Ladder Benchmark for LLM Cybersecurity Agents

Exploitation is not a binary event. It is a ladder of acquiring progressive capabilities, from executing a single buggy line of code to taking full control of the target. However, existing LLM security benchmarks treat a crash as exploitation success. That single binary outcome collapses the hard parts of exploitation: the transition from triggering a bug to constructing reusable primitives and control. We present ExploitBench, a capability-graded benchmark that decomposes exploitation into 16 measurable flags, from coverage and crash through sandbox primitives, arbitrary read/write, control-flow hijack, and arbitrary code execution. Each capability is verified by a deterministic oracle that uses a per-run randomized challenge-response for primitives, differential execution against ground-truth binaries to measure progress, and a signal-handler proof for code execution. We instantiate ExploitBench on 41 V8 bugs because V8 is both widely deployed and exploitation-hardened. We report three arms: <model,env> as the primary measurement of model-environment capability, <model,env, adaptive coaching> as a secondary arm that adds adaptive coaching to test whether targeted feedback shifts outcomes, and <model,env,harness> as an ablation that swaps in the model's native CLI to check whether vendor-side optimizations increase exploitation capabilities. Our results show a sharp capability split between publicly deployed frontier models and the private frontier. Across the 8 publicly deployed models tested, reaching the vulnerable code and triggering a crash is routine, but arbitrary code execution is not. The private model shows arbitrary code execution on approximately half. Overall, results suggest that exploit construction against hardened targets is an emerging frontier capability.

preprint2026arXiv

Solar Open Technical Report

We introduce Solar Open, a 102B-parameter bilingual Mixture-of-Experts language model for underserved languages. Solar Open demonstrates a systematic methodology for building competitive LLMs by addressing three interconnected challenges. First, to train effectively despite data scarcity for underserved languages, we synthesize 4.5T tokens of high-quality, domain-specific, and RL-oriented data. Second, we coordinate this data through a progressive curriculum jointly optimizing composition, quality thresholds, and domain coverage across 20 trillion tokens. Third, to enable reasoning capabilities through scalable RL, we apply our proposed framework SnapPO for efficient optimization. Across benchmarks in English and Korean, Solar Open achieves competitive performance, demonstrating the effectiveness of this methodology for underserved language AI development.

preprint2026arXiv

SPIO: Ensemble and Selective Strategies via LLM-Based Multi-Agent Planning in Automated Data Science

Large Language Models (LLMs) have enabled dynamic reasoning in automated data analytics, yet recent multi-agent systems remain limited by rigid, single-path workflows that restrict strategic exploration and often lead to suboptimal outcomes. To overcome these limitations, we propose SPIO (Sequential Plan Integration and Optimization), a framework that replaces rigid workflows with adaptive, multi-path planning across four core modules: data preprocessing, feature engineering, model selection, and hyperparameter tuning. In each module, specialized agents generate diverse candidate strategies, which are cascaded and refined by an optimization agent. SPIO offers two operating modes: SPIO-S for selecting a single optimal pipeline, and SPIO-E for ensembling top-k pipelines to maximize robustness. Extensive evaluations on Kaggle and OpenML benchmarks show that SPIO consistently outperforms state-of-the-art baselines, achieving an average performance gain of 5.6%. By explicitly exploring and integrating multiple solution paths, SPIO delivers a more flexible, accurate, and reliable foundation for automated data science.

preprint2025arXiv

Inverse-Designed Metasurfaces for Wavefront Restoration in Under-Display Camera Systems

Under-display camera (UDC) systems enable full-screen displays in smartphones by embedding the camera beneath the display panel, eliminating the need for notches or punch holes. However, the periodic pixel structures of display panels introduce significant optical diffraction effects, leading to imaging artifacts and degraded visual quality. Conventional approaches to mitigate these distortions, such as deep learning-based image reconstruction, are often computationally expensive and unsuitable for real-time applications in consumer electronics. This work introduces an inverse-designed metasurface for wavefront restoration, addressing diffraction-induced distortions without relying on external software processing. The proposed metasurface effectively suppresses higher-order diffraction modes caused by the metallic pixel structures, restores the optical wavefront, and enhances imaging quality across multiple wavelengths. By eliminating the need for software-based post-processing, our approach establishes a scalable, real-time optical solution for diffraction management in UDC systems. This advancement paves the way to achieve software-free real-time image restoration frameworks for many industrial applications.

preprint2023arXiv

Addressing Negative Transfer in Diffusion Models

Diffusion-based generative models have achieved remarkable success in various domains. It trains a shared model on denoising tasks that encompass different noise levels simultaneously, representing a form of multi-task learning (MTL). However, analyzing and improving diffusion models from an MTL perspective remains under-explored. In particular, MTL can sometimes lead to the well-known phenomenon of negative transfer, which results in the performance degradation of certain tasks due to conflicts between tasks. In this paper, we first aim to analyze diffusion training from an MTL standpoint, presenting two key observations: (O1) the task affinity between denoising tasks diminishes as the gap between noise levels widens, and (O2) negative transfer can arise even in diffusion training. Building upon these observations, we aim to enhance diffusion training by mitigating negative transfer. To achieve this, we propose leveraging existing MTL methods, but the presence of a huge number of denoising tasks makes this computationally expensive to calculate the necessary per-task loss or gradient. To address this challenge, we propose clustering the denoising tasks into small task clusters and applying MTL methods to them. Specifically, based on (O2), we employ interval clustering to enforce temporal proximity among denoising tasks within clusters. We show that interval clustering can be solved using dynamic programming, utilizing signal-to-noise ratio, timestep, and task affinity for clustering objectives. Through this, our approach addresses the issue of negative transfer in diffusion models by allowing for efficient computation of MTL methods. We validate the efficacy of proposed clustering and its integration with MTL methods through various experiments, demonstrating 1) improved generation quality and 2) faster training convergence of diffusion models.

preprint2022arXiv

CFA: Coupled-hypersphere-based Feature Adaptation for Target-Oriented Anomaly Localization

For a long time, anomaly localization has been widely used in industries. Previous studies focused on approximating the distribution of normal features without adaptation to a target dataset. However, since anomaly localization should precisely discriminate normal and abnormal features, the absence of adaptation may make the normality of abnormal features overestimated. Thus, we propose Coupled-hypersphere-based Feature Adaptation (CFA) which accomplishes sophisticated anomaly localization using features adapted to the target dataset. CFA consists of (1) a learnable patch descriptor that learns and embeds target-oriented features and (2) scalable memory bank independent of the size of the target dataset. And, CFA adopts transfer learning to increase the normal feature density so that abnormal features can be clearly distinguished by applying patch descriptor and memory bank to a pre-trained CNN. The proposed method outperforms the previous methods quantitatively and qualitatively. For example, it provides an AUROC score of 99.5% in anomaly detection and 98.5% in anomaly localization of MVTec AD benchmark. In addition, this paper points out the negative effects of biased features of pre-trained CNNs and emphasizes the importance of the adaptation to the target dataset. The code is publicly available at https://github.com/sungwool/CFA_for_anomaly_localization.

preprint2022arXiv

Ensemble Knowledge Guided Sub-network Search and Fine-tuning for Filter Pruning

Conventional NAS-based pruning algorithms aim to find the sub-network with the best validation performance. However, validation performance does not successfully represent test performance, i.e., potential performance. Also, although fine-tuning the pruned network to restore the performance drop is an inevitable process, few studies have handled this issue. This paper provides a novel Ensemble Knowledge Guidance (EKG) to solve both problems at once. First, we experimentally prove that the fluctuation of loss landscape can be an effective metric to evaluate the potential performance. In order to search a sub-network with the smoothest loss landscape at a low cost, we employ EKG as a search reward. EKG utilized for the following search iteration is composed of the ensemble knowledge of interim sub-networks, i.e., the by-products of the sub-network evaluation. Next, we reuse EKG to provide a gentle and informative guidance to the pruned network while fine-tuning the pruned network. Since EKG is implemented as a memory bank in both phases, it requires a negligible cost. For example, when pruning and training ResNet-50, just 315 GPU hours are required to remove around 45.04% of FLOPS without any performance degradation, which can operate even on a low-spec workstation. the implemented code is available at https://github.com/sseung0703/EKG.

preprint2022arXiv

FlexBlock: A Flexible DNN Training Accelerator with Multi-Mode Block Floating Point Support

Training deep neural networks (DNNs) is a computationally expensive job, which can take weeks or months even with high performance GPUs. As a remedy for this challenge, community has started exploring the use of more efficient data representations in the training process, e.g., block floating point (BFP). However, prior work on BFP-based DNN accelerators rely on a specific BFP representation making them less versatile. This paper builds upon an algorithmic observation that we can accelerate the training by leveraging multiple BFP precisions without compromising the finally achieved accuracy. Backed up by this algorithmic opportunity, we develop a flexible DNN training accelerator, dubbed FlexBlock, which supports three different BFP precision modes, possibly different among activation, weight, and gradient tensors. While several prior works proposed such multi-precision support for DNN accelerators, not only do they focus only on the inference, but also their core utilization is suboptimal at a fixed precision and specific layer types when the training is considered. Instead, FlexBlock is designed in such a way that high core utilization is achievable for i) various layer types, and ii) three BFP precisions by mapping data in a hierarchical manner to its compute units. We evaluate the effectiveness of FlexBlock architecture using well-known DNNs on CIFAR, ImageNet and WMT14 datasets. As a result, training in FlexBlock significantly improves the training speed by 1.5~5.3x and the energy efficiency by 2.4~7.0x on average compared to other training accelerators and incurs marginal accuracy loss compared to full-precision training.

preprint2020arXiv

Context-aware Dynamics Model for Generalization in Model-Based Reinforcement Learning

Model-based reinforcement learning (RL) enjoys several benefits, such as data-efficiency and planning, by learning a model of the environment's dynamics. However, learning a global model that can generalize across different dynamics is a challenging task. To tackle this problem, we decompose the task of learning a global dynamics model into two stages: (a) learning a context latent vector that captures the local dynamics, then (b) predicting the next state conditioned on it. In order to encode dynamics-specific information into the context latent vector, we introduce a novel loss function that encourages the context latent vector to be useful for predicting both forward and backward dynamics. The proposed method achieves superior generalization ability across various simulated robotics and control tasks, compared to existing RL schemes.

preprint2016arXiv

Evidence of electronic cloaking from chiral electron transport in bilayer graphene nanostructures

The coupling of charge carrier motion and pseudospin via chirality for massless Dirac fermions in monolayer graphene has generated dramatic consequences, such as the unusual quantum Hall effect and Klein tunneling. In bilayer graphene, charge carriers are massive Dirac fermions with a finite density of states at zero energy. Because of their non-relativistic nature, massive Dirac fermions can provide an even better test bed with which to clarify the importance of chirality in transport measurement than massless Dirac fermions in monolayer graphene. Here, we report an electronic cloaking effect as a manifestation of chirality by probing phase coherent transport in chemical-vapor-deposited bilayer graphene. Conductance oscillations with different periodicities were observed on extremely narrow bilayer graphene heterojunctions through electrostatic gating. Using a Fourier analysis technique, we identified the origin of each individual interference pattern. Importantly, the electron waves on the two sides of the potential barrier can be coupled through the evanescent waves inside the barrier, making the confined states underneath the barrier invisible to electrons. These findings provide direct evidence for the electronic cloaking effect and hold promise for the realization of pseudospintronics based on bilayer graphene.

preprint2016arXiv

Terahertz-visible two-photon rotational spectroscopy of cold OD-

We present a method to measure rotational transitions of molecular anions in the terahertz domain by sequential two-photon absorption. Ion excitation by bound-bound terahertz absorption is probed by absorption in the visible on a bound-free transition. The visible frequency is tuned to a state-selective photodetachment transition of the excited anions. This provides a terahertz action spectrum for just few hundred molecular ions. To demonstrate this we measure the two lowest rotational transitions, J=1<-0 and J =2<-1 of OD- anions in a cryogenic 22-pole trap. We obtain rotational transition frequencies of 598596.08(19) MHz for J=1<-0 and 1196791.57(27) MHz for J=2<-1 of OD-, in good agreement with their only previous measurement. This two-photon scheme opens up terahertz rovibrational spectroscopy for a range of molecular anions, in particular for polyatomic and cluster anions.

preprint2015arXiv

Cognitive Wireless Powered Network: Spectrum Sharing Models and Throughput Maximization

The recent advance in radio-frequency (RF) wireless energy transfer (WET) has motivated the study of wireless powered communication network (WPCN), in which distributed wireless devices are powered via dedicated WET by the hybrid access-point (H-AP) in the downlink (DL) for uplink (UL) wireless information transmission (WIT). In this paper, by exploiting the cognitive radio (CR) technique, we study a new type of CR enabled secondary WPCN, called cognitive WPCN, under spectrum sharing with the primary wireless communication system. In particular, we consider a cognitive WPCN, consisting of one single H-AP with constant power supply and distributed users, shares the same spectrum for its DL WET and UL WIT with an existing primary communication link, where the WPCN's WET/WIT and the primary link's WIT may interfere with each other. Under this new setup, we propose two coexisting models for spectrum sharing of the two systems, namely underlay and overlay based cognitive WPCNs, depending on different types of knowledge on the primary user transmission available at the cognitive WPCN. For each model, we maximize the sum-throughput of the cognitive WPCN by optimizing its transmission under different constraints applied to protect the primary user transmission. Analysis and simulation results are provided to compare the sum-throughput of the cognitive WPCN versus the achievable rate of the primary user in two coexisting models. It is shown that the overlay based cognitive WPCN outperforms the underlay based counterpart, thanks to its fully cooperative WET/WIT design with the primary WIT, while it also requires higher complexity for implementation.

preprint2015arXiv

Distributed Energy Beamforming with One-Bit Feedback

Energy beamforming (EB) is a key technique for achieving efficient radio-frequency (RF) transmission enabled wireless energy transfer (WET). By optimally designing the waveforms from multiple energy transmitters (ETs) over the wireless channels, they are constructively combined at the energy receiver (ER) to achieve an EB gain that scales with the number of ETs. However, the optimal design of transmit waveforms requires accurate channel state information (CSI) at the ETs, which is challenging to obtain in practical WET systems. In this paper, we propose a new channel training scheme to achieve optimal EB gain in a distributed WET system, where multiple separated ETs adjust their transmit phases to collaboratively send power to a single ER in an iterative manner, based on one-bit feedback from the ER per training interval which indicates the increase/decrease of the received power level from one particular ET over two preassigned transmit phases. The proposed EB algorithm can be efficiently implemented in practical WET systems even with a large number of distributed ETs, and is analytically shown to converge quickly to the optimal EB design as the number of feedback intervals per ET increases. Numerical results are provided to evaluate the performance of the proposed algorithm as compared to other distributed EB designs.

preprint2015arXiv

Rotational state-changing cold collisions of hydroxyl ions with helium

Cold molecules are important for many applications, from fundamental precision measurements, quantum information processing, quantum-controlled chemistry, to understanding the cold interstellar medium. Molecular ions are known to be cooled efficiently in sympathetic collisions with cold atoms or ions. However, little knowledge is available on the elementary cooling steps, because the determination of quantum state-to-state collision rates at low temperature is prohibitively challenging for both experiment and theory. Here we present a method to manipulate molecular quantum states by non-resonant photodetachment. Based on this we provide absolute quantum scattering rate coefficients under full quantum state control for the rotationally inelastic collision of hydroxyl anions with helium. Experiment and quantum scattering theory show excellent agreement without adjustable parameters. Very similar rate coefficients are obtained for two different isotopes, which is linked to several quantum scattering resonances appearing at different energies. The presented method is also applicable to polyatomic systems and will help shed light on non-radiative processes in polyaromatic hydrocarbons and protein chromophores.

preprint2014arXiv

Collaborative Wireless Energy and Information Transfer in Interference Channel

This paper studies the simultaneous wireless information and power transfer (SWIPT) in a multiuser wireless system, in which distributed transmitters send independent messages to their respective receivers, and at the same time cooperatively transmit wireless power to the receivers via energy beamforming. Accordingly, from the wireless information transmission (WIT) perspective, the system of interest can be modeled as the classic interference channel, while it also can be regarded as a distributed multiple-input multiple-output (MIMO) system for collaborative wireless energy transmission (WET). To enable both information decoding (ID) and energy harvesting (EH) in SWIPT, we adopt the low-complexity time switching operation at each receiver to switch between the ID and EH modes over scheduled time. Based on this hybrid model, we aim to characterize the achievable rate-energy (R-E) trade-offs in the multiuser SWIPT system under various transmitter-side collaboration schemes. Specifically, to facilitate the collaborative energy beamforming, we propose a new signal splitting scheme at the transmitters, where each transmit signal is generally composed of an information signal component and an energy signal component for WIT and WET, respectively. With this new scheme, first, we study the two-user SWIPT system and derive the optimal mode switching rule at the receivers and the corresponding transmit signal optimization to achieve various R-E trade-offs over the fading channel. We also compare the R-E performance of our proposed scheme with transmit energy beamforming and signal splitting against two existing schemes with partial or no cooperation of the transmitters, and show remarkable gains over these baseline schemes. Finally, the general case of SWIPT systems with more than two users is studied, for which we propose and compare two practical transmit collaboration schemes.

preprint2013arXiv

Measurement of collisions between rubidium atoms and optically dark rubidium ions

We measure the collision rate coefficient between laser cooled Rubidium (Rb) atoms in a magneto-optical trap (MOT) and optically dark Rb+ ions in an overlapping Paul trap. In such a mixture, the ions are created from the MOT atoms and allowed to accumulate in the ion trap, which results in a significant reduction in the number of steady state MOT atoms. A theoretical rate equation model is developed to describe the evolution of the MOT atom number, due to ionization and ion-atom collision, and derive an expression for the ion-atom collision rate coefficient. The loss of MOT atoms is studied systematically, by sequentially switching on the various mechanisms in the experiment. Combining the measurements with the model allows the direct determination of the ion-atom collision rate coefficient. Finally the scope of the experimental technique developed here is discussed.

preprint2013arXiv

Opportunistic Wireless Energy Harvesting in Cognitive Radio Networks

Wireless networks can be self-sustaining by harvesting energy from ambient radio-frequency (RF) signals. Recently, researchers have made progress on designing efficient circuits and devices for RF energy harvesting suitable for low-power wireless applications. Motivated by this and building upon the classic cognitive radio (CR) network model, this paper proposes a novel method for wireless networks coexisting where low-power mobiles in a secondary network, called secondary transmitters (STs), harvest ambient RF energy from transmissions by nearby active transmitters in a primary network, called primary transmitters (PTs), while opportunistically accessing the spectrum licensed to the primary network. We consider a stochastic-geometry model in which PTs and STs are distributed as independent homogeneous Poisson point processes (HPPPs) and communicate with their intended receivers at fixed distances. Each PT is associated with a guard zone to protect its intended receiver from ST's interference, and at the same time delivers RF energy to STs located in its harvesting zone. Based on the proposed model, we analyze the transmission probability of STs and the resulting spatial throughput of the secondary network. The optimal transmission power and density of STs are derived for maximizing the secondary network throughput under the given outage-probability constraints in the two coexisting networks, which reveal key insights to the optimal network design. Finally, we show that our analytical result can be generally applied to a non-CR setup, where distributed wireless power chargers are deployed to power coexisting wireless transmitters in a sensor network.

preprint2012arXiv

Cognitive Energy Harvesting and Transmission from a Network Perspective

Wireless networks can be self-sustaining by harvesting energy from radio-frequency (RF) signals. Building on classic cognitive radio networks, we propose a novel method for network coexisting where mobiles from a secondary network, called secondary transmitters (STs), either harvest energy from transmissions by nearby transmitters from a primary network, called primary transmitters (PTs), or transmit information if PTs are sufficiently far away; STs store harvested energy in rechargeable batteries with finite capacity and use all available energy for subsequent transmission when batteries are fully charged. In this model, each PT is centered at a guard zone and a harvesting zone that are disks with given radiuses; a ST harvests energy if it lies in some harvesting zone, transmits fixed-power signals if it is outside all guard zones or else idles. Based on this model, the spatial throughput of the secondary network is maximized using a stochastic-geometry model where PTs and STs are modeled as independent homogeneous Poisson point processes (HPPPs), under the outage constraints for coexisting networks and obtained in a simple closed-form. It is observed from the result that the maximum secondary throughput decreases linearly with the growing PT density, and the optimal ST density is inversely proportional to the derived transmission probability for STs.

preprint2012arXiv

Coverage and Economy of Cellular Networks with Many Base Stations

The performance of a cellular network can be significantly improved by employing many base stations (BSs), which shortens transmission distances. However, there exist no known results on quantifying the performance gains from deploying many BSs. To address this issue, we adopt a stochastic-geometry model of the downlink cellular network and analyze the mobile outage probability. Specifically, given Poisson distributed BSs, the outage probability is shown to diminish inversely with the increasing ratio between the BS and mobile densities. Furthermore, we analyze the optimal tradeoff between the performance gain from increasing the BS density and the resultant network cost accounting for energy consumption, BS hardware and backhaul cables. The optimal BS density is proved to be proportional to the square root of the mobile density and the inverse of the square root of the cost factors considered.

preprint2012arXiv

Evidence for Extraction of Photoexcited Hot Carriers from Graphene

We report evidence of nonequilibrium hot carriers extraction from graphene by gate-dependent photocurrent study. Scanning photocurrent excited by femtosecond pulse laser shows unusual gate dependence compared with continuous wave (CW) laser excitation. Power dependence studies further confirm that the photocarriers extracted at the metal/graphene contact are nonequilibrium hot carriers. Hot carrier extraction is found to be most efficient near the Dirac point where carrier lifetime reaches maximum. These observations not only provide evidence of hot carrier extraction from graphene, but also open the door for graphene based hot carrier optoelectronics.

preprint2012arXiv

Flexible and Transparent All-Graphene Circuits for Quaternary Digital Modulations

In modern communication system, modulation is a key function that embeds the baseband signal (information) into a carrier wave so that it can be successfully broadcasted through a medium such as air or cables. A flexible signal modulation scheme is hence essential to wide range of applications based on flexible electronics. Here we report a fully bendable all-graphene modulator circuit with the capability to encode a carrier signal with quaternary digital information for the first time. By exploiting the ambipolarity and the nonlinearity in a graphene transistor, we demonstrated two types of quaternary modulation schemes: 4-ary amplitude-shift keying (4-ASK) and quadrature phase-shift keying (QPSK). Remarkably, 4-ASK and QPSK can be realized with just 1 and 2 all-graphene transistors, respectively, representing a drastic reduction in circuit complexity when compared with conventional digital modulators. In addition, the circuit is not only flexible but also highly transparent (~95% transmittance) owing to their all-graphene design with every component (channel, interconnects, load resistor, and source/drain/gate electrodes) fabricated from graphene films. Taken together, these results represent a significant step toward achieving a high speed communication system that can be monolithically integrated on a flexible and transparent platform.

preprint2011arXiv

Combined ion and atom trap for low temperature ion-atom physics

We report an experimental apparatus and technique which simultaneously traps ions and cold atoms with spatial overlap. Such an apparatus is motivated by the study of ion-atom processes at temperatures ranging from hot to ultra-cold. This area is a largely unexplored domain of physics with cold trapped atoms. In this article we discuss the general design considerations for combining these two traps and present our experimental setup. The ion trap and atom traps are characterized independently of each other. The simultaneous operation of both is then described and experimental signatures of the effect of the ions and cold-atoms on each other are presented. In conclusion the use of such an instrument for several problems in physics and chemistry is briefly discussed.

preprint2011arXiv

Homogeneous bilayer graphene film based flexible transparent conductor

Graphene is considered a promising candidate to replace conventional transparent conductors due to its low opacity, high carrier mobility and flexible structure. Multi-layer graphene or stacked single layer graphenes have been investigated in the past but both have their drawbacks. The uniformity of multi-layer graphene is still questionable, and single layer graphene stacks require many transfer processes to achieve sufficiently low sheet resistance. In this work, bilayer graphene film grown with low pressure chemical vapor deposition was used as a transparent conductor for the first time. The technique was demonstrated to be highly efficient in fabricating a conductive and uniform transparent conductor compared to multi-layer or single layer graphene. Four transfers of bilayer graphene yielded a transparent conducting film with a sheet resistance of 180 Ω_{\square} at a transmittance of 83%. In addition, bilayer graphene films transferred onto plastic substrate showed remarkable robustness against bending, with sheet resistance change less than 15% at 2.14% strain, a 20-fold improvement over commercial indium oxide films.

preprint2011arXiv

Sympathetic and swap cooling of trapped ions by cold atoms in a MOT

A mixed system of cooled and trapped, ions and atoms, paves the way for ion assisted cold chemistry and novel many body studies. Due to the different individual trapping mechanisms, trapped atoms are significantly colder than trapped ions, therefore in the combined system, the strong binary ion$-$atom interaction results in heat flow from ions to atoms. Conversely, trapped ions can also get collisionally heated by the cold atoms, making the resulting equilibrium between ions and atoms intriguing. Here we experimentally demonstrate, Rubidium ions (Rb$^+$) cool in contact with magneto-optically trapped (MOT) Rb atoms, contrary to the general expectation of ion heating for equal ion and atom masses. The cooling mechanism is explained theoretically and substantiated with numerical simulations. The importance of resonant charge exchange (RCx) collisions, which allows swap cooling of ions with atoms, wherein a single glancing collision event brings a fast ion to rest, is discussed.

preprint2010arXiv

Wafer Scale Homogeneous Bilayer Graphene Films by Chemical Vapor Deposition

The discovery of electric field induced bandgap opening in bilayer graphene opens new door for making semiconducting graphene without aggressive size scaling or using expensive substrates. However, bilayer graphene samples have been limited to um size scale thus far, and synthesis of wafer scale bilayer graphene posts tremendous challenge. Here we report homogeneous bilayer graphene films over at least 2 inch x 2 inch area, synthesized by chemical vapor deposition on copper foil and subsequently transferred to arbitrary substrates. The bilayer nature of graphene film is verified by Raman spectroscopy, atomic force microscopy (AFM), and transmission electron microscopy (TEM). Importantly, spatially resolved Raman spectroscopy confirms a bilayer coverage of over 99%. The homogeneity of the film is further supported by electrical transport measurements on dual-gate bilayer graphene transistors, in which bandgap opening is observed in 98% of the devices.

preprint2009arXiv

A Three Dimensional Lattice of Ion Traps

We propose an ion trap configuration such that individual traps can be stacked together in a three dimensional simple cubic arrangement. The isolated trap as well as the extended array of ion traps are characterized for different locations in the lattice, illustrating the robustness of the lattice of traps concept. Ease in the addressing of ions at each lattice site, individually or simultaneously, makes this system naturally suitable for a number of experiments. Application of this trap to precision spectroscopy, quantum information processing and the study of few particle interacting system are discussed.

Seunghyun Lee

What is connected

Connect this record

See the researcher in context

Building this map preview

27 published item(s)

Automated Visualization Code Synthesis via Multi-Path Reasoning and Feedback-Driven Optimization

ExploitBench: A Capability Ladder Benchmark for LLM Cybersecurity Agents

Solar Open Technical Report

SPIO: Ensemble and Selective Strategies via LLM-Based Multi-Agent Planning in Automated Data Science

Inverse-Designed Metasurfaces for Wavefront Restoration in Under-Display Camera Systems

Addressing Negative Transfer in Diffusion Models

CFA: Coupled-hypersphere-based Feature Adaptation for Target-Oriented Anomaly Localization

Ensemble Knowledge Guided Sub-network Search and Fine-tuning for Filter Pruning

FlexBlock: A Flexible DNN Training Accelerator with Multi-Mode Block Floating Point Support

Context-aware Dynamics Model for Generalization in Model-Based Reinforcement Learning

Evidence of electronic cloaking from chiral electron transport in bilayer graphene nanostructures

Terahertz-visible two-photon rotational spectroscopy of cold OD-

Cognitive Wireless Powered Network: Spectrum Sharing Models and Throughput Maximization

Distributed Energy Beamforming with One-Bit Feedback

Rotational state-changing cold collisions of hydroxyl ions with helium

Collaborative Wireless Energy and Information Transfer in Interference Channel

Measurement of collisions between rubidium atoms and optically dark rubidium ions

Opportunistic Wireless Energy Harvesting in Cognitive Radio Networks

Cognitive Energy Harvesting and Transmission from a Network Perspective

Coverage and Economy of Cellular Networks with Many Base Stations

Evidence for Extraction of Photoexcited Hot Carriers from Graphene

Flexible and Transparent All-Graphene Circuits for Quaternary Digital Modulations

Combined ion and atom trap for low temperature ion-atom physics

Homogeneous bilayer graphene film based flexible transparent conductor

Sympathetic and swap cooling of trapped ions by cold atoms in a MOT

Wafer Scale Homogeneous Bilayer Graphene Films by Chemical Vapor Deposition

A Three Dimensional Lattice of Ion Traps