Searcharxiv⌕ Search

arXiv subjects

Minseok Kim

Publications and source records attributed to Minseok Kim.

At least 19 recordsLinked to original sources
arXiv2026

Partially Observable Zero-shot coordination by Predicting Intention of Partner

Zero-shot coordination in embodied settings requires acting while the partner is intermittently out of view, leaving existing methods with ambiguous partner representations and uncertainty over hidden partner states. We propose Predicting Intention of Partner (PIP) to jointly address these challenges. PIP uses a Joint-view VAE to distill richer training-time evidence from the union of both agents' local observations into a partner representation available from local observations alone. Partner-state Belief networks further infer the partner's hidden location and behavioral tendencies from the ego agent's interaction history. We evaluate PIP in Burrito-PO, Overcooked-PO, and a Melting Pot substrate, together with a human evaluation in Burrito-PO. PIP attains the highest mean performance among the compared methods across all three benchmarks. Human evaluation and diagnostic analyses further support coordination with unseen partners and the contributions of both components under partner occlusion.

cs.AI↗
arXiv2026

Where Does the Audio Jailbreak Live? A Controlled Frequency-Depth Audit of AdvWave-P on Qwen2-Audio

We audit frequency and decoder-depth claims for AdvWave-P, an additive audio jailbreak, on Qwen2-Audio. The protocol masks frequency components of the perturbation in the short-time Fourier transform (STFT) domain and measures attack success and audio-span representations. On 520 AdvBench prompts, the primary judge labels 76.7% of adversarial inputs as jailbreaks. A condition-blind, single-annotator validation yields a Rogan-Gladen sensitivity estimate of 0.87 for this condition (about 0.83-0.95 with validation-rate uncertainty); this correction is not applied to masked conditions. The apparent frequency ranking depends on the partition: energy share alone predicts the standard eight-band ranking (Spearman's rho = 0.95), and equal-Hz and equal-energy partitions show that masking any tested band can sharply reduce attack success. At a finer 16-band equal-energy resolution, however, masking the narrow 7520-7960 Hz band leaves ASR at 0.10, which remains unresolved without a matched control. Matched-energy scattered-removal tests show that contiguous removal is more damaging in lower bands, while both forms approach the floor in upper bands. Global rescaling leaves ASR near baseline but tests amplitude sensitivity rather than frequency location. In a re-optimization pilot (n = 20), tested single- and two-band supports reach ASRs of 0.00, 0.25, and 0.40, while random supports covering about half the STFT bins reach a mean of 0.86. In a prompt- and energy-adjusted model, audio-span divergence is associated with band necessity, with the coefficient rising from +0.63 at the projector output to +0.93 at layer 30 (contrast +0.293, 95% CI [0.11, 0.51]). This association is not a causal localization, and single-layer patching does not establish a causal layer. The results support partition-aware auditing of frequency claims, leaving the fine-resolution top-band result and broader generality open.

cs.SD↗
arXiv2026

Full-Wave-Calibrated Element-Wise RIS Modeling With Cross-Aperture Coefficient Transfer for Multipath Channel Prediction

Practical reconfigurable intelligent surfaces (RISs) can exhibit deterministic parasitic scattering that is not captured by idealized element-wise models. As a result, such models may overestimate the gain of the intended RIS-assisted path and bias multipath prediction. This paper develops a full-wave-calibrated element-wise model using three Bragg-order basis functions to represent the intended and dominant parasitic scattering components. Environmental multipath is incorporated by identifying RIS--Rx reflection sequences with ray tracing (RT) and unfolding them by image theory into path-dependent image points. This allows direct and reflected RIS-assisted paths to be evaluated using the same per-element kernel and coherently combined with Tx--Rx bypass paths. To reduce the full-wave calibration burden for large RISs under a prescribed focusing configuration, only the three Bragg-order coefficients are transferred from a 25-by-25 calibration aperture, while the target-aperture basis functions and geometry are recomputed. At 154 GHz, the transferred coefficients keep the intended-order errors within 0.8 dB for the 50-by-50 and 100-by-100 RISs, while reducing the full-wave calibration time for the prescribed configuration from 7.56 to 1.08 h relative to direct 100-by-100 calibration. For multipath validation with the 50-by-50 RIS, the calibrated model using the same transferred coefficients keeps the nominal-Rx gain error within 0.88 dB across four PEC reflector configurations, compared with 1.75--5.51 dB for the uncalibrated general model, and reduces the error from 6.49 to 0.23 dB in the scaled indoor environment.

eess.SP↗
arXiv2026

Monotone Neural Policy Iteration for High-Dimensional First-Order Hamilton--Jacobi--Bellman Equations

We analyze a neural semi-discrete method for high-dimensional first-order Hamilton-Jacobi-Bellman (HJB) equations with known or learned dynamics. Centered differences and an artificial viscosity $Nh=O(h)$ define a monotone operator evaluated through $2d+1$ shifted network queries; policy iteration solves the resulting Bellman equation without a tensor grid. At fixed $h$, the sharp componentwise condition $\max_i|f_i|\le2N$ turns every frozen-policy operator into a nearest-neighbor Markov-chain generator with a policy-independent total jump rate. Uniformization gives whole-space well-posedness for measurable feedbacks, an explicit Poisson-tail bound on the numerical domain of dependence, and boundary-free localization. The representation also yields a posteriori policy-evaluation bounds that account for residual and learned-model errors. A greedy-gap analysis controls inexact policy iteration at fixed $h$; a separate consistency estimate connects the semi-discrete equation to the continuous HJB equation. Experiments reproduce the extremal tail, show rates consistent with $O(\sqrt h)$ and nearly $h$-independent exact-policy-iteration decay, and assess empirical estimator effectivity. A nonsmooth example shows that the continuous residual can miss a non-viscosity solution, whereas the shifted residual detects the defect. Further tests provide a structured interval-verified certificate calibration, an early-budget benefit of policy freezing for bang-bang control, and learned-dynamics diagnostics. A structured nonlinear problem with active compact-control constraints is tested against a manufactured semi-discrete reference through $d=1024$.

cs.LG↗
arXiv2026

FaVOR: LLM-Based Agentic Framework for Factor Mining via Empirical Validation

Traditional finance relies on experts to hand-craft factors through a principled process grounded in economic rationale. Recent LLM-based multi-agent systems have automated this process, scaling factor mining far beyond manual effort. However, these automated approaches optimize directly for returns and rarely check whether a generated factor still expresses the economic hypothesis that motivated it. We identify this inconsistency between mathematical form and economic meaning as a structural failure mode of return-oriented automation. The resulting factors blur the line between real signals and spurious correlations and break down across regime shifts. We propose FaVOR (Factor Validation through Observable Reasoning), an agentic framework that restructures factor mining around hypothesis-level evidence rather than return outcomes. In place of the standard hypothesis-to-formula leap, FaVOR enforces a three-stage consistency loop tying mathematical form to economic rationale throughout. (1) Decomposition splits a broad economic hypothesis into independent observable conditions. (2) Validation checks whether each factor reflects its intended condition. (3) Integration merges them into a composite whose structure remains interpretable. On the CSI 500 and S&P 500 in 2025, FaVOR outperforms existing baselines while remaining effective across regimes. FaVOR shows that hypothesis-grounded factor discovery produces signals that are interpretable by construction, regime-robust, and economically faithful. The code is available at https://github.com/damilab/FaVOR.

cs.AI↗
arXiv2026

Finite-temperature spin diffusion in the two-dimensional XY model

We present a combined theory-experiment study to quantify spin diffusion in the square lattice quantum spin-1/2 XY model at finite temperature. On the theory side, we leverage a recently developed dynamical high-temperature expansion method to faithfully capture the long spatiotemporal scales of the hydrodynamic regime. Experimental results are obtained from an optical lattice hard-core boson quantum simulator. The excellent agreement of spin diffusion constants marks a breakthrough in spin-transport beyond one dimension and for the quantitative validation of state-of-the-art quantum simulation platforms. We also provide theory predictions for future experiments on dynamic spin conductivity or anisotropy-induced integrability breaking.

cond-mat.quant-gas↗
arXiv2026

Real-time feedback control of ELM frequency using divertor gas puffing and its effects on tungsten-induced radiation and plasma performance in KSTAR

The edge-localized mode (ELM) frequency ($f_{\mathrm{ELM}}$) was successfully controlled in real time on KSTAR using a proportional-integral (PI) feedback controller, employing a $\mathrm{D}_2$ divertor gas puff as the actuator under tungsten lower-divertor conditions. The controller accurately tracked a two-step target---a 30 Hz increase in $f_{\mathrm{ELM}}$ for 4 s, followed by a 30 Hz decrease for 3 s---yielding mean and median absolute percentage errors of approximately 13% and 12%, respectively. Compared to a reference discharge, the actively controlled shot did not exhibit a significant drop in volume-integrated core radiation, confirming that excessive gas use merely degrades overall plasma performance. However, when contrasted with the exponential increase in core radiation observed in the absence of divertor gas puffing, these results underscore the critical need for real-time optimization. Specifically, divertor gas commands must be actively managed to maintain an $f_{\mathrm{ELM}}$ sufficient for flushing tungsten from the core while maximizing global plasma performance.

physics.plasm-ph↗
arXiv2026

Synthesized-Isotropic Narrowband Channel Parameter Extraction from Angle-Resolved Wideband Channel Measurements

Angle-resolved channel sounding using antenna arrays or mechanically steered high-gain antennas is widely employed at millimeter-wave and terahertz bands. To extract antenna-independent large-scale channel parameters such as path loss, delay spread, and angular spread, the radiation-pattern effects embedded in the measured responses must be properly compensated. This paper revisits the technical challenges of path loss/path gain calculation from angle-resolved wideband measurements, with emphasis on angular-domain power integration where the scan beams are inherently non-orthogonal and simple power summation leads to biased isotropic-equivalent power estimates. We first formulate the synthesized-isotropic narrowband power in a unified matrix form and introduce a beam-accumulation correction factor, including an offset-averaged variant to mitigate scalloping due to off-grid angles. The proposed framework is validated through simulations using channel models and 154~GHz corridor measurements.

eess.SP↗
arXiv2026

Physics-Informed Policy Iteration for High-Dimensional Hamilton--Jacobi--Bellman Equations: Interior Error Bounds without Boundary Data

We develop a physics-informed policy-iteration method for stationary second-order Hamilton--Jacobi--Bellman equations arising in continuous-time stochastic control. Each policy-evaluation step is a linear elliptic PDE and is approximated by a mesh-free neural residual solver; policy improvement is then performed pointwise from the surrogate gradient. The analysis addresses bounded-domain training without prescribed boundary data. We prove well-posedness of the PDE-defined evaluation for every Borel Markov policy, establish Lipschitz stability of the greedy map on bounded gradient ranges, and derive an exponential attenuation estimate for unresolved boundary information. These ingredients yield a closed finite-step interior error bound whose floors are determined by a continuous $L^p$ residual, with a finite exponent $p>d$, and an attenuated amplitude term. The experiments measure the quantities in the estimate. On a linear--quadratic testbed with an exact reference, the gradient-error floor scales nearly linearly with a fresh-sample $L^p$ residual estimate, and finite-difference probes identify when a training buffer is beneficial. At matched architecture and budget, linear fixed-policy training becomes markedly more reliable than direct minimization of the nonlinear HJB residual as the tested problems become more difficult. The method also produces effective feedback on an inverted pendulum, a planar quadrotor, and a 100-dimensional posterior-seeking problem. These nonlinear tests expose two limitations not visible from sampled residuals alone: the collocation distribution may miss the region visited by the learned closed loop, and evaluation without boundary data may leave undetermined the gradient component on which the greedy update depends. On-policy collocation, rollout-anchored evaluation, and rollout-based stopping provide effective model-only safeguards.

cs.LG↗
arXiv2026

Urban Macro/Microcellular Channel Characterization at 4.85 GHz With Literature-Referenced Upper FR1-to-FR3 Cross-Band Analysis

The transition from 5G to 6G requires frequency-dependent, physically consistent radio channel models across the upper-FR1/FR3 transition region, particularly in the under-explored $4$--$8$~GHz region targeted in the current WRC-$27$ studies, where outdoor urban channel measurements and characterizations remain scarce. This paper presents a $4.85$~GHz measurement-anchored study of urban channels and a literature-referenced cross-band analysis. Double-directional measurements were conducted at $4.85$~GHz in urban macrocell (UMa) and urban microcell (UMi) routes in Yokohama, Japan, from which path loss, delay spread (DS), azimuth spread of arrival/departure (ASA/ASD), $K$-factor, and route-dependent spatial-consistency statistics were extracted. To align these results in a broader cross-band context, the measured $4.85$~GHz large-scale parameter (LSP) means were combined with scenario-matched literature anchors to derive log-log trends for DS, ASA, and ASD over an approximately $4$--$28$~GHz range around the $7.125$~GHz upper-FR1/FR3 cross-band boundary. The resulting trends were compared with 3GPP UMa/UMi reference parameterizations over the same interval, and the sensitivity of the UMi DS fit was examined via leave-one-out analysis. Because the cross-band analysis still relies on a single in-house measurement band alongside heterogeneous anchors from different campaigns, it is presented as measurement-informed and indicative rather than as a definitive multi-band model. The paper therefore contributes both a detailed, parameterized $4.85$~GHz urban measurement reference and a bounded literature-referenced view of channel behavior near the upper-FR1/FR3 transition.

eess.SP↗
arXiv2026

$p$-twisted Selmer near-companion curves

Let $E_1$ and $E_2$ be elliptic curves over a number field $K$. In \cite{scc}, Mazur and Rubin define the concept of $n$-Selmer near-companions and conjecture that if $E_1$ and $E_2$ are $n$-Selmer near-companions over $K$, then $E_1[n]$ is $G_K$-isomorphic to $E_2[n]$. Yu proves the conjecture on $n$-Selmer near-companion curves in the case $n=2$. We we introduce the notion of $p$-twisted Selmer near-companions ($p$-TSNC) over $K$ and prove that if $E_1$ and $E_2$ are $p$-TSNC over $K$, then $K(E_1[p])=K(E_2[p])$ under certain conditions.

math.NT↗
arXiv2026

Multimodal Distribution Matching for Vision-Language Dataset Distillation

Dataset distillation compresses large training sets into compact synthetic datasets while preserving downstream performance. As modern systems increasingly operate on paired vision-language inputs, multimodal distillation must preserve representation quality and cross-modal alignment under tight compute and memory budgets, yet prior methods often require heavy computes and overlook their correlations. To address this, we present Multimodal Distribution Matching (MDM), a geometry-aware framework for efficient and generalizable multimodal distillation. Specifically, MDM integrates complementary components at the data, model, and loss levels. At the data level, it initializes synthetic image-text pairs by sampling from clusters in the joint embedding space. At the model level, it forms a mixed teacher by interpolating independently fine-tuned models in weight space according to their angular deviation from the pretrained anchor. At the loss level, it matches joint distributions on the unit hypersphere using a geometry-aware matching objective that exploits the joint features in the cross-modal agreement and discrepancy directions along with symmetric contrastive learning. Across image-text retrieval benchmarks with cross-architecture evaluation, MDM yields compact synthetic sets that preserve multimodal semantics, substantially reduce distillation cost, and remain robust across architectures.

cs.CV↗
arXiv2026

Environment-Aware Channel Prediction for Vehicular Communications: A Multimodal Visual Feature Fusion Framework

The deep integration of communication with intelligence and sensing, as a defining vision of 6G, renders environment-aware channel prediction a key enabling technology. As a representative 6G application, vehicular communications require accurate and forward-looking channel prediction under stringent reliability, latency, and adaptability demands. Traditional empirical and deterministic models remain limited in balancing accuracy, generalization, and deployability, while the growing availability of onboard and roadside sensing devices offers a promising source of environmental priors. This paper proposes an environment-aware channel prediction framework based on multimodal visual feature fusion. Using GPS data and vehicle-side panoramic RGB images, together with semantic segmentation and depth estimation, the framework extracts semantic, depth, and position features through a three-branch architecture and performs adaptive multimodal fusion via a squeeze-excitation attention gating module. For 360-dimensional angular power spectrum (APS) prediction, a dedicated regression head and a composite multi-constraint loss are further designed. As a result, joint prediction of path loss (PL), delay spread (DS), azimuth spread of arrival (ASA), azimuth spread of departure (ASD), and APS is achieved. Experiments on a synchronized urban V2I measurement dataset yield the best root mean square error (RMSE) of 3.26 dB for PL, RMSEs of 37.66 ns, 5.05 degrees, and 5.08 degrees for DS, ASA, and ASD, respectively, and mean/median APS cosine similarities of 0.9342/0.9571, demonstrating strong accuracy, generalization, and practical potential for intelligent channel prediction in 6G vehicular communications.

cs.CV↗
arXiv2026

HarassGuard: Detecting Harassment Behaviors in Social Virtual Reality with Vision-Language Models

Social Virtual Reality (VR) platforms provide immersive social experiences but also expose users to serious risks of online harassment. Existing safety measures are largely reactive, while proactive solutions that detect harassment behavior during an incident often depend on sensitive biometric data, raising privacy concerns. In this paper, we present HarassGuard, a vision-language model (VLM) based system that detects physical harassment in social VR using only visual input. We construct an IRB-approved harassment vision dataset, apply prompt engineering, and fine-tune VLMs to detect harassment behavior by considering contextual information in social VR. Experimental results demonstrate that HarassGuard achieves competitive performance compared to state-of-the-art baselines (i.e., LSTM/CNN, Transformer), reaching an accuracy of up to 88.09% in binary classification and 68.85% in multi-class classification. Notably, HarassGuard matches these baselines while using significantly fewer fine-tuning samples (200 vs. 1,115), offering unique advantages in contextual reasoning and privacy-preserving detection.

cs.CV↗
arXiv2026

Aligning Paralinguistic Understanding and Generation in Speech LLMs via Multi-Task Reinforcement Learning

Speech large language models (LLMs) observe paralinguistic cues such as prosody, emotion, and non-verbal sounds--crucial for intent understanding. However, leveraging these cues faces challenges: limited training data, annotation difficulty, and models exploiting lexical shortcuts over paralinguistic signals. We propose multi-task reinforcement learning (RL) with chain-of-thought prompting that elicits explicit affective reasoning. To address data scarcity, we introduce a paralinguistics-aware speech LLM (PALLM) that jointly optimizes sentiment classification from audio and paralinguistics-aware response generation via a two-stage pipeline. Experiments demonstrate that our approach improves paralinguistics understanding over both supervised baselines and strong proprietary models (Gemini-2.5-Pro, GPT-4o-audio) by 8-12% on Expresso, IEMOCAP, and RAVDESS. The results show that modeling paralinguistic reasoning with multi-task RL is crucial for building emotionally intelligent speech LLMs.

cs.CL↗
arXiv2026

A Physics-Informed, Global-in-Time Neural Particle Method for the Spatially Homogeneous Landau Equation

We propose a physics-informed neural particle method (PINN--PM) for the spatially homogeneous Landau equation. The method adopts a Lagrangian interacting-particle formulation and jointly parameterizes the time-dependent score and the characteristic flow map with neural networks. Instead of advancing particles through explicit time stepping, the Landau dynamics is enforced via a continuous-time residual defined along particle trajectories. This design removes time-discretization error and yields a mesh-free solver that can be queried at arbitrary times without sequential integration. We establish a rigorous stability analysis in an $L^2_v$ framework. The deviation between learned and exact characteristics is controlled by three interpretable sources: (i) score approximation error, (ii) empirical particle approximation error, and (iii) the physics residual of the neural flow. This trajectory estimate propagates to density reconstruction, where we derive an $L^2_v$ error bound for kernel density estimators combining classical bias--variance terms with a trajectory-induced contribution. Using Hyvarinen's identity, we further relate the oracle score-matching gap to the $L^2_v$ score error and show that the empirical loss concentrates at the Monte Carlo rate, yielding computable a posteriori accuracy certificates. Numerical experiments on analytical benchmarks, including the two- and three-dimensional BKW solutions, as well as reference-free configurations, demonstrate stable transport, preservation of macroscopic invariants, and competitive or improved accuracy compared with time-stepping score-based particle and blob methods while using significantly fewer particles.

math.NA↗
arXiv2026

Loss-Optimized Reconfigurable Nonlocal Metasurface-aided Cavity Antenna

This paper presents the design and experimental demonstration of a reconfigurable cavity excited nonlocal metasurface antenna capable of wide angle dynamic beam steering. The antenna is synthesized using a volume surface integral equation based framework that rigorously captures nonlocal mutual coupling among metasurface unit cells. To ensure physical consistency, the numerically characterized resistance and reactance relationship of the tunable unit cells is directly incorporated into the synthesis, enabling precise far-field synthesis while minimizing Ohmic losses. The proposed approach is applied to a 10 GHz cavity fed metasurface antenna composed of 24 independently controlled varactor-loaded unit cells. Numerical simulations and near-field measurements demonstrate stable beam steering with a range of 80 degrees across broadside with excellent agreement between measured and simulated radiation patterns. These results confirm the effectiveness of the proposed framework for the realization of compact, reconfigurable cavity-excited metasurface antennas.

physics.app-ph↗
arXiv2026

Quasi-Deterministic Modeling of Sub-THz Band Access Channels in Street Canyon Environments

Sub-terahertz (sub-THz) frequencies (100--300 GHz) are expected to play a key role in beyond-5G and 6G mobile networks. However, their quasi-optical propagation characteristics require new channel models beyond sub-100 GHz extrapolations. This paper presents an extensive double-directional (D-D) channel measurement campaign conducted in an outdoor street-canyon environment at 154 GHz and 300 GHz under both line-of-sight (LoS) and non-line-of-sight (NLoS) conditions using an in-house-developed multi-tone frequency-domain channel sounder. Based on these measurements, clustering with merged datasets across the two frequencies enables comparative analyses that identify both common and distinct multipath clusters, as well as the frequency dependence of cluster-level characteristics. A quasi-deterministic (Q-D) channel model is then proposed, combining deterministic components, such as LoS and single-bounce reflections from side walls, with random components. Large-scale parameters (path loss, delay spread, angular spread, and Rician $K$-factor) are also evaluated. These results provide valuable insights into sub-THz propagation in urban street canyons and contribute toward the development of accurate, channel models for future 6G systems.

eess.SP↗