Researcher profile

Sunwoo Lee

Sunwoo Lee contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

Machine Learning Artificial Intelligence astro-ph.IM Multiagent Systems Distributed, Parallel, and Cluster Computing Performance physics.ed-ph

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author

9works

0followers

7topics

4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

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

preprint2026arXiv

Enabling Weak Client Participation via On-device Knowledge Distillation in Heterogeneous Federated Learning

Online Knowledge Distillation (KD) is recently highlighted to train large models in Federated Learning (FL) environments. Many existing studies adopt the logit ensemble method to perform KD on the server side. However, they often assume that unlabeled data collected at the edge is centralized on the server. Moreover, the logit ensemble method personalizes local models, which can degrade the quality of soft targets, especially when data is highly non-IID. To address these critical limitations,we propose a novel on-device KD-based heterogeneous FL method. Our approach leverages a small auxiliary model to learn from labeled local data. Subsequently, a subset of clients with strong system resources transfers knowledge to a large model through on-device KD using their unlabeled data. Our extensive experiments demonstrate that our on-device KD-based heterogeneous FL method effectively utilizes the system resources of all edge devices as well as the unlabeled data, resulting in higher accuracy compared to SOTA KD-based FL methods.

preprint2026arXiv

Ghosted Layers: Unconstrained Activation Alignment for Recovering Layer-Pruned LLMs

Layer pruning removes entire Transformer decoder blocks from large language models, but introduces a mismatch between the hidden state received by the next surviving layer and the distribution it was trained to process, leading to significant performance degradation. We propose Ghosted Layers, a training-free recovery module that addresses this issue by solving a boundary activation alignment problem. Our method derives a closed-form optimal linear operator from a small calibration set to reconstruct the activation discrepancy introduced by the pruned layers. We show that this solution corresponds to the unconstrained optimum of the alignment objective, whereas existing methods are restricted to constrained solutions over limited operator subspaces. Experiments across multiple LLM backbones and pruning strategies demonstrate that our method consistently improves accuracy and perplexity over prior training-free baselines, while preserving the efficiency gains of layer pruning.

preprint2026arXiv

Interaction-Breaking Adversarial Learning Framework for Robust Multi-Agent Reinforcement Learning

Cooperation is central to multi-agent reinforcement learning (MARL), yet learned coordination can be fragile when external perturbations disrupt inter-agent interactions. Prior robust MARL methods have primarily considered value-oriented attacks, leaving a gap in robustness when interaction structures themselves are corrupted. In this paper, we propose an interaction-breaking adversarial learning (IBAL) framework that takes an information-theoretic view to construct attacks that impede coordination by perturbing agents' observations and actions, and trains agents to perform reliably under such disruptions. Empirically, our approach improves robustness over existing robust MARL baselines across diverse attack settings and yields stronger performance even under agent-missing scenarios.

preprint2026arXiv

Shaping Zero-Shot Coordination via State Blocking

Zero-shot coordination (ZSC) aims to enable agents to cooperate with independently trained partners without prior interaction, a key requirement for real-world multi-agent systems and human-AI collaboration. Existing approaches have largely emphasized increasing partner diversity during training, yet such strategies often fall short of achieving reliable generalization to unseen partners. We introduce State-Blocked Coordination (SBC), a simple yet effective framework that improves ZSC by inducing diverse interaction scenarios without direct environment modification. Specifically, SBC generates a family of virtual environments through state blocking, allowing agents to experience a wide range of suboptimal partner policies. Across multiple benchmarks, SBC demonstrates superior performance in zero-shot coordination, including strong generalization to human partners.

preprint2022arXiv

Layer-wise Adaptive Model Aggregation for Scalable Federated Learning

In Federated Learning, a common approach for aggregating local models across clients is periodic averaging of the full model parameters. It is, however, known that different layers of neural networks can have a different degree of model discrepancy across the clients. The conventional full aggregation scheme does not consider such a difference and synchronizes the whole model parameters at once, resulting in inefficient network bandwidth consumption. Aggregating the parameters that are similar across the clients does not make meaningful training progress while increasing the communication cost. We propose FedLAMA, a layer-wise model aggregation scheme for scalable Federated Learning. FedLAMA adaptively adjusts the aggregation interval in a layer-wise manner, jointly considering the model discrepancy and the communication cost. The layer-wise aggregation method enables to finely control the aggregation interval to relax the aggregation frequency without a significant impact on the model accuracy. Our empirical study shows that FedLAMA reduces the communication cost by up to 60% for IID data and 70% for non-IID data while achieving a comparable accuracy to FedAvg.

preprint2022arXiv

Partial Model Averaging in Federated Learning: Performance Guarantees and Benefits

Local Stochastic Gradient Descent (SGD) with periodic model averaging (FedAvg) is a foundational algorithm in Federated Learning. The algorithm independently runs SGD on multiple workers and periodically averages the model across all the workers. When local SGD runs with many workers, however, the periodic averaging causes a significant model discrepancy across the workers making the global loss converge slowly. While recent advanced optimization methods tackle the issue focused on non-IID settings, there still exists the model discrepancy issue due to the underlying periodic model averaging. We propose a partial model averaging framework that mitigates the model discrepancy issue in Federated Learning. The partial averaging encourages the local models to stay close to each other on parameter space, and it enables to more effectively minimize the global loss. Given a fixed number of iterations and a large number of workers (128), the partial averaging achieves up to 2.2% higher validation accuracy than the periodic full averaging.

preprint2020arXiv

Development of Linear Astigmatism Free -- Three Mirror System (LAF-TMS)

We present the development of Linear Astigmatism Free - Three Mirror System (LAF-TMS). This is a prototype of an off-axis telescope that enables very wide field of view (FoV) infrared satellites that can observe Paschen-$α$ emission, zodiacal light, integrated star light, and other infrared sources. It has the entrance pupil diameter of 150 mm, the focal length of 500 mm, and the FoV of 5.5$^\circ$ $\times$ 4.1$^\circ$. LAF-TMS is an obscuration-free off-axis system with minimal out-of-field baffling and no optical support structure diffraction. This optical design is analytically optimized to remove linear astigmatism and to reduce high-order aberrations. Sensitivity analysis and Monte-Carlo simulation reveal that tilt errors are the most sensitive alignment parameters that allow $\sim$1$^\prime$. Optomechanical structure accurately mounts aluminum mirrors, and withstands satellite-level vibration environments. LAF-TMS shows optical performance with 37 $μ$m FWHM of the point source image satisfying Nyquist sampling requirements for typical 18 $μ$m pixel Infrared array detectors. The surface figure errors of mirrors and scattered light from the tertiary mirror with 4.9 nm surface micro roughness may affect the measured point spread function (PSF). Optical tests successfully demonstrate constant optical performance over wide FoV, indicating that LAF-TMS suppresses linear astigmatism and high-order aberrations.

preprint2020arXiv

Transformable Reflective Telescope for optical testing and education

We propose and experimentally demonstrate the Transformable Reflective Telescope (TRT) Kit for educational purposes and for performing various optical tests with a single kit. The TRT Kit is a portable optical bench setup suitable for interferometry, spectroscopy, measuring stray light, and developing adaptive optics, among other uses. Supplementary modules may be integrated easily thanks to the modular design of the TRT Kit. The Kit consists of five units; a primary mirror module, a secondary mirror module, a mounting base module, a baffle module, and an alignment module. Precise alignment and focusing are achieved using a precision optical rail on the alignment module. The TRT Kit transforms into three telescope configurations: Newtonian, Cassegrain, and Gregorian. Students change telescope configurations by exchanging the secondary mirror. The portable design and the aluminum primary mirror of the TRT Kit enable students to perform experiments in various environments. The minimized baffle design utilizes commercial telescope tubes, allowing users to look directly into the optical system while suppressing stray light down to $\sim$10$^{-8}$ point source transmittance (PST). The TRT Kit was tested using a point source and field images. Point source measurement of the Newtonian telescope configuration resulted in an 80\% encircled energy diameter (EED) of 23.8 $μ$m.

preprint2019arXiv

Improving MPI Collective I/O Performance With Intra-node Request Aggregation

Two-phase I/O is a well-known strategy for implementing collective MPI-IO functions. It redistributes I/O requests among the calling processes into a form that minimizes the file access costs. As modern parallel computers continue to grow into the exascale era, the communication cost of such request redistribution can quickly overwhelm collective I/O performance. This effect has been observed from parallel jobs that run on multiple compute nodes with a high count of MPI processes on each node. To reduce the communication cost, we present a new design for collective I/O by adding an extra communication layer that performs request aggregation among processes within the same compute nodes. This approach can significantly reduce inter-node communication congestion when redistributing the I/O requests. We evaluate the performance and compare with the original two-phase I/O on a Cray XC40 parallel computer with Intel KNL processors. Using I/O patterns from two large-scale production applications and an I/O benchmark, we show the performance improvement of up to 29 times when running 16384 MPI processes on 256 compute nodes.

Institution

Affiliation not imported yet

This author record came from a source that does not expose affiliation metadata. Once the author claims the profile or we enrich the record from another provider, this section will link to the concrete institution.

Topic footprint