arXiv subjects

Min-Gyu Kim

Publications and source records attributed to Min-Gyu Kim.

11 recordsLinked to original sources

arXiv2026

The Verbose Context Problem in Medical Records

The verbose context problem occurs when structured concepts have token-inefficient textual representations. This bottleneck is acute in population health: cohort-level analysis of longitudinal patient records requires reasoning over thousands of medically-coded events, often exceeding 400K tokens in total. We present PopMedQA, a benchmark isolating this problem through computational tasks on groups of longitudinal patient records. We construct the benchmark using neopatient, a new library for language-controlled generation of artificial patient records. Through extensive ablations -- including prompting strategies, prompt compression, and agentic decomposition -- we find that domain-independent methods fail to alleviate the verbose context problem. There remains significant opportunity to exploit domain-specific structure in language model inputs for population-scale reasoning.

cs.CL↗

arXiv2026

ECG-Reasoning-Benchmark: A Benchmark for Evaluating Clinical Reasoning Capabilities in ECG Interpretation

While Multimodal Large Language Models (MLLMs) show promising performance in automated electrocardiogram interpretation, it remains unclear whether they genuinely perform actual step-by-step reasoning or just rely on superficial visual cues. To investigate this, we introduce \textbf{ECG-Reasoning-Benchmark}, a novel multi-turn evaluation framework comprising over 6,400 samples to systematically assess step-by-step reasoning across 17 core ECG diagnoses. Our comprehensive evaluation of state-of-the-art models reveals a critical failure in executing multi-step logical deduction. Although models possess the medical knowledge to retrieve clinical criteria for a diagnosis, they exhibit near-zero success rates (6% Completion) in maintaining a complete reasoning chain, primarily failing to ground the corresponding ECG findings to the actual visual evidence in the ECG signal. These results demonstrate that current MLLMs bypass actual visual interpretation, exposing a critical flaw in existing training paradigms and underscoring the necessity for robust, reasoning-centric medical AI. The code and data are available at https://github.com/Jwoo5/ecg-reasoning-benchmark.

cs.LG↗

arXiv2026

ECG-Agent: On-Device Tool-Calling Agent for ECG Multi-Turn Dialogue

Recent advances in Multimodal Large Language Models have rapidly expanded to electrocardiograms, focusing on classification, report generation, and single-turn QA tasks. However, these models fall short in real-world scenarios, lacking multi-turn conversational ability, on-device efficiency, and precise understanding of ECG measurements such as the PQRST intervals. To address these limitations, we introduce ECG-Agent, the first LLM-based tool-calling agent for multi-turn ECG dialogue. To facilitate its development and evaluation, we also present ECG-Multi-Turn-Dialogue (ECG-MTD) dataset, a collection of realistic user-assistant multi-turn dialogues for diverse ECG lead configurations. We develop ECG-Agents in various sizes, from on-device capable to larger agents. Experimental results show that ECG-Agents outperform baseline ECG-LLMs in response accuracy. Furthermore, on-device agents achieve comparable performance to larger agents in various evaluations that assess response accuracy, tool-calling ability, and hallucinations, demonstrating their viability for real-world applications.

cs.AI↗

arXiv2025

Dynamic Policy Learning for Legged Robot with Simplified Model Pretraining and Model-Homotopy-Inspired Transfer

Generating dynamic motions for legged robots remains a challenging problem. While reinforcement learning has achieved notable success in various legged locomotion tasks, producing highly dynamic behaviors often requires extensive reward tuning or high-quality demonstrations. Leveraging reduced-order models can help mitigate these challenges. However, the model discrepancy poses a significant challenge when transferring policies to full-body dynamics environments. In this work, we introduce a continuation-based learning framework that combines simplified model pretraining and model-homotopy-inspired transfer to efficiently generate and refine complex dynamic behaviors. First, we pretrain the policy using a single rigid body model to capture core motion patterns in a simplified environment. Next, we employ a continuation strategy to progressively transfer the policy to the full-body environment, minimizing performance loss. To define the continuation path, we introduce a parametric transition path from the single rigid body model to the full-body model by gradually redistributing mass and inertia between the trunk and legs. The proposed method achieves faster convergence and demonstrates superior stability during the transfer process compared to baseline methods. Our framework is validated on a range of dynamic tasks, including flips and wall-assisted maneuvers, and is successfully deployed on a real quadrupedal robot.

cs.RO↗

arXiv2025

Dynamically-Consistent Trajectory Optimization for Legged Robots via Contact Point Decomposition

To generate reliable motion for legged robots through trajectory optimization, it is crucial to simultaneously compute the robot's path and contact sequence, as well as accurately consider the dynamics in the problem formulation. In this paper, we present a phase-based trajectory optimization that ensures the feasibility of translational dynamics and friction cone constraints throughout the entire trajectory. Specifically, our approach leverages the superposition properties of linear differential equations to decouple the translational dynamics for each contact point, which operates under different phase sequences. Furthermore, we utilize the differentiation matrix of B{\'e}zier polynomials to derive an analytical relationship between the robot's position and force, thereby ensuring the consistent satisfaction of translational dynamics. Additionally, by exploiting the convex closure property of B{\'e}zier polynomials, our method ensures compliance with friction cone constraints. Using the aforementioned approach, the proposed trajectory optimization framework can generate dynamically reliable motions with various gait sequences for legged robots. We validate our framework using a quadruped robot model, focusing on the feasibility of dynamics and motion generation. The source code and additional materials are publicly available at our project page: https://sangmin11.github.io/Analytical_Dynamics_TO_Project/.

cs.RO↗

arXiv2025

A Modular Residual Learning Framework to Enhance Model-Based Approach for Robust Locomotion

This paper presents a novel approach that combines the advantages of both model-based and learning-based frameworks to achieve robust locomotion. The residual modules are integrated with each corresponding part of the model-based framework, a footstep planner and dynamic model designed using heuristics, to complement performance degradation caused by a model mismatch. By utilizing a modular structure and selecting the appropriate learning-based method for each residual module, our framework demonstrates improved control performance in environments with high uncertainty, while also achieving higher learning efficiency compared to baseline methods. Moreover, we observed that our proposed methodology not only enhances control performance but also provides additional benefits, such as making nominal controllers more robust to parameter tuning. To investigate the feasibility of our framework, we demonstrated residual modules combined with model predictive control in a real quadrupedal robot. Despite uncertainties beyond the simulation, the robot successfully maintains balance and tracks the commanded velocity.

cs.RO↗

arXiv2025

How Balanced Should Causal Covariates Be?

Covariate balancing is a popular technique for controlling confounding in observational studies. It finds weights for the treatment group which are close to uniform, but make the group's covariate means (approximately) equal to those of the entire sample. A crucial question is: how approximate should the balancing be, in order to minimize the error of the final estimate? Current guidance is derived from heuristic or asymptotic analyses, which are uninformative when the size of the sample is small compared to the number of covariates. This paper presents the first rigorous, nonasymptotic analysis of covariate balancing; specifically, we use PAC-Bayesian techniques to derive valid, finite-sample confidence intervals for the treatment effect. More generally, we prove these guarantees for a flexible form of covariate balancing where the regularization parameters weighting the tradeoff between bias (imbalance) and variance (divergence from uniform) are optimized, not fixed. This gives rise to a new balancing algorithm which empirically delivers superior adaptivity. Our overall contribution is to make covariate balancing a more reliable method for causal inference.

stat.ME↗

arXiv2025

Online Friction Coefficient Identification for Legged Robots on Slippery Terrain Using Smoothed Contact Gradients

This paper proposes an online friction coefficient identification framework for legged robots on slippery terrain. The approach formulates the optimization problem to minimize the sum of residuals between actual and predicted states parameterized by the friction coefficient in rigid body contact dynamics. Notably, the proposed framework leverages the analytic smoothed gradient of contact impulses, obtained by smoothing the complementarity condition of Coulomb friction, to solve the issue of non-informative gradients induced from the nonsmooth contact dynamics. Moreover, we introduce the rejection method to filter out data with high normal contact velocity following contact initiations during friction coefficient identification for legged robots. To validate the proposed framework, we conduct the experiments using a quadrupedal robot platform, KAIST HOUND, on slippery and nonslippery terrain. We observe that our framework achieves fast and consistent friction coefficient identification within various initial conditions.

cs.RO↗

arXiv2025

Tracking Error Based Fault Tolerant Scheme for Marine Vehicles with Thruster Redundancy

This paper proposes an active model-based fault and failure tolerant control scheme for a class of marine vehicles with thruster redundancy. Unlike widely used state and parameter estimation methods, where the estimation errors are utilized to generate residual, in this paper we directly apply the trajectory tracking error terms to construct residual and detect thruster fault and failure in the steady state of the tracking system. As for identification or diagnosis, this paper proposes a novel scheme through a detailed examination of the tracking error trends and the combinations of thruster configurations. Since this fault detection and identification operates within the same closed-loop of the tracking control system, control reconfiguration can be easily achieved by adjusting the weight parameter of the isolated thruster to minimize tracking errors or residual. Numerical studies with the real world vehicle model is also carried out to verify the effectiveness of the proposed method.

eess.SY↗

arXiv2021

Doping effects on the valence bond solid of Li$_{2}$RuO$_{3}$ with Mn substitution

$Li_{2}RuO_{3}$ with a honeycomb structure undergoes a drastic transition from a regular honeycomb lattice with the $C2/m$ space group to a valence bond solid state of the $P2_{1}/m$ space group with an extremely strong dimerization at 550 K. We synthesized $Li_{2}Ru_{1-x}Mn_{x}O_{3}$ with a full solid solution and investigated doping effects on the valence bond solid state as a function of Mn content. The valence bond solid state is found to be stable up to $x = 0.2$, based on our extensive experiments: structural studies, resistivity, and magnetic susceptibility. On the other hand, the extended x-ray absorption fine structure analyses show that the dimer local structure remains robust even above $x = 0.2$ with a minimal effect on the dimer bond length. This indicates that the locally-disordered dimer structure survives well into the Mn-rich phase even though the thermodynamically stable average structure has the $C2/m$ space group. Our results prove that the dimer formation in $Li_{2}RuO_{3}$ is predominantly a local phenomenon driven by the formation of orbitally-assisted metal-metal bonds and that these dimers are relatively robust against doping-induced disorder.

cond-mat.str-el↗

arXiv2011

Measuring Light from the Epoch of Reionization with CIBER, the Cosmic Infrared Background Experiment

Ultraviolet emission from the first generation of stars in the Universe ionized the intergalactic medium in a process which was completed by z~6; the wavelength of these photons has been redshifted by (1+z) into the near infrared today and can be measured using instruments situated above the Earth's atmosphere. First flying in February 2009, the Cosmic Infrared Background Experiment (CIBER) comprises four instruments housed in a single reusable sounding rocket borne payload. CIBER will measure spatial anisotropies in the extragalactic IR background caused by cosmological structure from the epoch of reionization using two broadband imaging instruments, make a detailed characterization of the spectral shape of the IR background using a low resolution spectrometer, and measure the absolute brightness of the Zodical light foreground with a high resolution spectrometer in each of our six science fields. This paper presents the scientific motivation for CIBER and details of its first two flights, including a review of the published scientific results from the first flight and an outlook for future reionization science with CIBER data.

astro-ph.CO↗