arXiv subjects

Minkyoung Kim

Publications and source records attributed to Minkyoung Kim.

10 recordsLinked to original sources

arXiv2026

Chance-constrained selection of sequential intervention strategies from counterfactual estimates

Many operational decisions are sequences of interventions under a cumulative resource limit, such as a maintenance schedule within a crew-hour budget. Choosing among them calls for the outcome and the cumulative cost each would produce, counterfactual quantities identified from observational data. Two strategies with the same expected cost can exceed the budget at very different rates, so constraining the mean does not bound how often an overrun occurs. Prior two-step architectures, recently extended to continuous doses, constrain the mean cost rather than its tail and allocate at a single decision point. Methods that do bound a cost tail take its distribution from a specified model rather than identifying it from data. We present a predict-then-optimize framework. In the prediction step, any estimator returning an outcome value and a cost distribution supplies what the decision rule consumes, so the predictor is interchangeable. In the optimization step, a chance-constrained selection over a finite candidate set bounds the probability that the cumulative cost exceeds the budget. That tail does not decompose across stages, so each strategy is scored whole. Sweeping the tolerated violation probability traces a safety-utility frontier, and distribution-free finite-sample bounds cover violation and outcome shortfall. Four of five environments, spanning clinical treatment and equipment maintenance, supply exact counterfactual ground truth; the fifth carries real outcomes from a digital-health micro-randomized trial. Across them, the rule holds the budget where a point-estimate rule overruns it, at an outcome cost the frontier makes explicit. All code is available at https://github.com/mfriendly/counterfactual-chance-selection

stat.ME↗

arXiv2026

Bilevel Graph Structure Learning, Revisited: Inner-Channel Origins of the Reported Gain

Bilevel graph structure learning is widely understood to improve graph neural networks by jointly optimizing model parameters and a learned graph structure, with the resulting performance gain attributed to the rewired adjacency. We find that this attribution may be overstated: training-dynamics effects in the inner loop, rather than the rewiring itself, capture a substantial share of the gain. To establish this, we introduce frozen-$\phi$, a control that freezes the graph while retaining the inner-loop training schedule. This decomposes the bilevel gain into an inner channel of $T$-step training dynamics with implicit gradient regularization and a graph channel of the graph rewiring itself. On spatio-temporal flow forecasting the inner channel matches or exceeds the full bilevel pipeline, accounting for 78-101% of the gain; on node classification it accounts for 37-44% under a Bernoulli edge-level parameterization. We also verify that classical spectral diagnostics can dissociate from task gain. We propose frozen-$\phi$ as a standardized diagnostic for bilevel graph structure learning, with graph distillation as a method-agnostic complement. A three-precondition framework further predicts the sign of the bilevel gain on all six benchmarks.

cs.LG↗

arXiv2025

LIVE-GS: Online LiDAR-Inertial-Visual State Estimation and Globally Consistent Mapping with 3D Gaussian Splatting

While 3D Gaussian Splatting (3DGS) enabled photorealistic mapping, its integration into SLAM has largely followed traditional camera-centric pipelines. As a result, they inherit well-known weaknesses such as high computational load, failure in texture-poor or illumination-varying environments, and limited operational range, particularly for RGB-D setups. On the other hand, LiDAR emerges as a robust alternative, but its integration with 3DGS introduces new challenges, such as the need for tighter global alignment for photorealistic quality and prolonged optimization times caused by sparse data. To address these challenges, we propose LIVE-GS, an online LiDAR-Inertial Visual SLAM framework that tightly couples 3D Gaussian Splatting with LiDAR-based surfels to ensure high-precision map consistency through global geometric optimization. Particularly, to handle sparse data, our system employs a depth-invariant Gaussian initialization strategy for efficient representation and a bounded sigmoid constraint to prevent uncontrolled Gaussian growth. Experiments on public and our datasets demonstrate competitive performance in rendering quality and map-building efficiency compared with representative 3DGS SLAM baselines.

cs.RO↗

arXiv2024

Mitigating Adversarial Attacks in LLMs through Defensive Suffix Generation

Large language models (LLMs) have exhibited outstanding performance in natural language processing tasks. However, these models remain susceptible to adversarial attacks in which slight input perturbations can lead to harmful or misleading outputs. A gradient-based defensive suffix generation algorithm is designed to bolster the robustness of LLMs. By appending carefully optimized defensive suffixes to input prompts, the algorithm mitigates adversarial influences while preserving the models' utility. To enhance adversarial understanding, a novel total loss function ($L_{\text{total}}$) combining defensive loss ($L_{\text{def}}$) and adversarial loss ($L_{\text{adv}}$) generates defensive suffixes more effectively. Experimental evaluations conducted on open-source LLMs such as Gemma-7B, mistral-7B, Llama2-7B, and Llama2-13B show that the proposed method reduces attack success rates (ASR) by an average of 11\% compared to models without defensive suffixes. Additionally, the perplexity score of Gemma-7B decreased from 6.57 to 3.93 when applying the defensive suffix generated by openELM-270M. Furthermore, TruthfulQA evaluations demonstrate consistent improvements with Truthfulness scores increasing by up to 10\% across tested configurations. This approach significantly enhances the security of LLMs in critical applications without requiring extensive retraining.

cs.CV↗

arXiv2024

Enhancing Clinical Efficiency through LLM: Discharge Note Generation for Cardiac Patients

Medical documentation, including discharge notes, is crucial for ensuring patient care quality, continuity, and effective medical communication. However, the manual creation of these documents is not only time-consuming but also prone to inconsistencies and potential errors. The automation of this documentation process using artificial intelligence (AI) represents a promising area of innovation in healthcare. This study directly addresses the inefficiencies and inaccuracies in creating discharge notes manually, particularly for cardiac patients, by employing AI techniques, specifically large language model (LLM). Utilizing a substantial dataset from a cardiology center, encompassing wide-ranging medical records and physician assessments, our research evaluates the capability of LLM to enhance the documentation process. Among the various models assessed, Mistral-7B distinguished itself by accurately generating discharge notes that significantly improve both documentation efficiency and the continuity of care for patients. These notes underwent rigorous qualitative evaluation by medical expert, receiving high marks for their clinical relevance, completeness, readability, and contribution to informed decision-making and care planning. Coupled with quantitative analyses, these results confirm Mistral-7B's efficacy in distilling complex medical information into concise, coherent summaries. Overall, our findings illuminate the considerable promise of specialized LLM, such as Mistral-7B, in refining healthcare documentation workflows and advancing patient care. This study lays the groundwork for further integrating advanced AI technologies in healthcare, demonstrating their potential to revolutionize patient documentation and support better care outcomes.

cs.CL↗

arXiv2024

HyperCLOVA X Technical Report

We introduce HyperCLOVA X, a family of large language models (LLMs) tailored to the Korean language and culture, along with competitive capabilities in English, math, and coding. HyperCLOVA X was trained on a balanced mix of Korean, English, and code data, followed by instruction-tuning with high-quality human-annotated datasets while abiding by strict safety guidelines reflecting our commitment to responsible AI. The model is evaluated across various benchmarks, including comprehensive reasoning, knowledge, commonsense, factuality, coding, math, chatting, instruction-following, and harmlessness, in both Korean and English. HyperCLOVA X exhibits strong reasoning capabilities in Korean backed by a deep understanding of the language and cultural nuances. Further analysis of the inherent bilingual nature and its extension to multilingualism highlights the model's cross-lingual proficiency and strong generalization ability to untargeted languages, including machine translation between several language pairs and cross-lingual inference tasks. We believe that HyperCLOVA X can provide helpful guidance for regions or countries in developing their sovereign LLMs.

cs.CL↗

arXiv2018

Causal Inference in Disease Spread across a Heterogeneous Social System

Diffusion processes are governed by external triggers and internal dynamics in complex systems. Timely and cost-effective control of infectious disease spread critically relies on uncovering the underlying diffusion mechanisms, which is challenging due to invisible causality between events and their time-evolving intensity. We infer causal relationships between infections and quantify the reflexivity of a meta-population, the level of feedback on event occurrences by its internal dynamics (likelihood of a regional outbreak triggered by previous cases). These are enabled by our new proposed model, the Latent Influence Point Process (LIPP) which models disease spread by incorporating macro-level internal dynamics of meta-populations based on human mobility. We analyse 15-year dengue cases in Queensland, Australia. From our causal inference, outbreaks are more likely driven by statewide global diffusion over time, leading to complex behavior of disease spread. In terms of reflexivity, precursory growth and symmetric decline in populous regions is attributed to slow but persistent feedback on preceding outbreaks via inter-group dynamics, while abrupt growth but sharp decline in peripheral areas is led by rapid but inconstant feedback via intra-group dynamics. Our proposed model reveals probabilistic causal relationships between discrete events based on intra- and inter-group dynamics and also covers direct and indirect diffusion processes (contact-based and vector-borne disease transmissions).

q-bio.PE↗

arXiv2017

Modeling Reflexivity of Social Systems in Disease Spread

Diffusion processes in a social system are governed by external triggers and internal excitations via interactions between individuals over social networks. Underlying mechanisms are crucial to understand emergent phenomena in the real world and accordingly establish effective strategies. However, it is challenging to reveal the dynamics of a target diffusion process due to invisible causality between events and their time-evolving intensity. In this study, we propose the Latent Influence Point Process model (LIPP) by incorporating external heterogeneity and internal dynamics of meta-populations based on human mobility. Our proposed model quantifies the reflexivity of a social system, which is the level of feedback on event occurrences by its internal dynamics. As an exemplary case study, we investigate dengue outbreaks in Queensland, Australia during the last 15 years. We find that abrupt and steady growth of disease outbreaks relate to exogenous and endogenous influences respectively. Similar diffusion trends between regions reflect synchronous reflexivity of a regional social system, likely driven by human mobility.

cs.SI↗

arXiv2017

Predictability of Irregular Human Mobility

Understanding human mobility is critical for decision support in areas from urban planning to infectious diseases control. Prior work has focused on tracking daily logs of outdoor mobility without considering relevant context, which contain a mixture of regular and irregular human movement for a range of purposes, and thus diverse effects on the dynamics have been ignored. This study aims to focus on irregular human movement of different meta-populations with various purposes. We propose approaches to estimate the predictability of mobility in different contexts. With our survey data from international and domestic visitors to Australia, we found that the travel patterns of Europeans visiting for holidays are less predictable than those visiting for education, while East Asian visitors show the opposite patterns, ie, more predictable for holidays than for education. Domestic residents from the most populous Australian states exhibit the most unpredictable patterns, while visitors from less populated states show the highest predictable movement.

physics.soc-ph↗

arXiv2017

Universal Components of Real-world Diffusion Dynamics based on Point Processes

Bursts in human and natural activities are highly clustered in time, suggesting that these activities are influenced by previous events within the social or natural system. Bursty behavior in the real world conveys information of underlying diffusion processes, which have been the focus of diverse scientific communities from online social media to criminology and epidemiology. However, universal components of real-world diffusion dynamics that cut across disciplines remain unexplored. Here, we introduce a wide range of diffusion processes across disciplines and propose universal components of diffusion frameworks. We apply these components to diffusion-based studies of human disease spread, through a case study of the vector-borne disease dengue. The proposed universality of diffusion can motivate transdisciplinary research and provide a fundamental framework for diffusion models.

cs.SI↗