arXiv subjects

Yang Hong

Publications and source records attributed to Yang Hong.

At least 19 recordsLinked to original sources

arXiv2026

Same Request, Different Boundary: Evaluating Cybersecurity Assistance across Conversational Contexts

Large Language Models (LLMs) can solve complex problems, but their misuse in high-risk domains can lead to severe consequences. Model providers therefore restrict assistance for potentially harmful requests. Refusing all cybersecurity requests would therefore harm legitimate users. Providers need a mechanism to block malicious use without denying legitimate assistance to defenders. Existing cybersecurity-specific datasets evaluate this mechanism, but none considers the conversational context of a request. We introduce 3R-Bench (Refusal, Repetition, and Revision), a benchmark of 150 real-world cybersecurity requests augmented with two adversarial conversational settings, and evaluate eight LLMs on it. Prior assistant behavior strongly changes responses to an unchanged request: among 376 available pairs from a 400-pair panel, compliance rises from 62.0% after refused history to 85.1% after accepted history. The opposite pattern appears under dialogue decomposition. In comparison, compliance falls from 501/800 direct responses to 172/800 after dialogue; among 738 pairs returning model-authored text in both conditions, the decrease is 45.1 points. Failure feedback recovers only a small fraction of this loss.

cs.AI↗

arXiv2026

SoK: When Safe Agents Fail Together: The Security of Multi Agent LLM Systems

Safe agents can fail together. Multi-agent LLM systems (MAS) move information, state, decisions, and authority across principal boundaries, creating failures that local checks may miss. Without an execution-level view, a multi-agent setting can easily be mistaken for evidence of a genuinely multi-agent security effect. We thus systematize MAS security through an execution-centered analysis of 197 works, covering six interaction interfaces, four adversary positions, seven system-level risks, and eight recurring attack paths. We introduce an A-I-R framework that organizes attacks by adversary position, interaction interface, and resulting system-level risk, unifying otherwise fragmented attack mechanisms across MAS. We organize defenses through a five-part contract covering path target, observation, intervention, trust boundary, and recovery, and identify path closure and recovery as key challenges. We audit 44 evaluation and benchmark works and identify open challenges in isolating interaction effects, designing comparable and diagnostic metrics, supporting reuse across MAS designs, and evaluating open-system operation. Together, these findings motivate an interaction-aware view of MAS security: trace attacks end to end, test whether defenses close those paths, and evaluate system-level effects with appropriate counterfactuals.

cs.CR↗

arXiv2026

'OpenBloom': A Stigma-Sensitive LLM Design Probe for Navigating Reproductive Well-being Conversations with Young Adults

The growing use of large language models (LLMs) by young adults seeking sensitive health information has raised important questions in Human-AI Interaction about how these systems can support understanding and navigation of reproductive well-being. In response to Feminist HCI principles, we introduce OpenBloom, a web application and an exploratory design probe that uses LLMs to generate question-based prompts from reproductive health articles. Through a user study with 34 young adults across 136 interactions with OpenBloom, we provide an initial assessment of the system while exploring how participants' reflections engage with culture and value sensitivities. We found that while OpenBloom outputs meet expectations of "safe" and non-offensive, they tend to paraphrase or rely on factual recall, which may lead to value dilution. We discuss implications under contestability and value-sensitive frameworks for future LLM-mediated reproductive health technologies and towards responsible AI discourse.

cs.HC↗

arXiv2026

Why does AI unlock new possibilities in STEM education? A Bibliometric Analysis of Trends and Future Agenda

STEM education faces challenges in personalization and interdisciplinary integration. AI technology has brought new possibilities, but the mechanisms by which AI reshapes the STEM education ecosystem require systematic investigation. This study employs bibliometric methods to analyze 242 publications from 2015-2025, constructing knowledge maps to reveal the evolutionary trajectory. The findings show that the field has transformed from intelligent tutoring systems to inquiry-based learning and computational thinking cultivation driven by LLMs. AI's key contribution lies in providing intelligent scaffolding that lowers the threshold for understanding knowledge. In this sense, AI is a core driving force promoting its shift from knowledge transmission to capability development.

cs.CY↗

arXiv2026

HydroAgent: Closing the Gap Between Frontier LLMs and Human Experts in Hydrologic Model Calibration via Simulator-Grounded RL

Calibrating distributed hydrologic models is a critical bottleneck across operational water resources management - streamflow prediction, reservoir operation, drought monitoring, infrastructure design, and flood forecasting all depend on it. Each basin demands an expert to translate hydrograph signatures into adjustments of a high-dimensional parameter vector, and the resulting workflow does not transfer between watersheds. We ask: can frontier large language model (LLM) agents replace the human hydrologic modeler, and if not, what would it take? We benchmark nine frontier LLM agents - Claude Opus 4.6/4.7, Sonnet 4.6, GPT-5/5.4/5.4-pro, and Gemini 2.5-pro/3.1-pro/3-flash - on the operational CREST distributed hydrologic model used by the U.S. National Weather Service for flash-flood forecasting. Best-of-twenty-rounds Nash-Sutcliffe Efficiency (NSE) across four held-out gauges spanning 329-40,792 km2 ranges from -0.16 (GPT-5.4) to 0.75 (Sonnet 4.6); the ceiling reproduces across all three vendors and capability tiers, with the strongest models concentrating in the 0.65-0.75 band, and no model reaches the human-expert reference except Opus-4.7 on one gauge. We argue this gap is not a parameter-count problem but a domain-grounding problem. We then propose HYDROAGENT, fine-tuning open-weight Qwen3-4B with supervised fine-tuning on 2,576 expert calibration trajectories and Group-Relative Policy Optimization using NSE as a verifiable reward from online CREST simulations - reinforcement learning with simulation feedback (RLSF). For Earth system science, a small domain-tuned policy with simulator-in-the-loop RL is a more compute-efficient and physically faithful path than scaling generic frontier models, and the multi-modal richness of Earth data - remote sensing, in-situ time series, and forecaster narrative - makes domain agents a leveraged direction for AI in physical science.

cs.LG↗

arXiv2026

Cloud-top infrared observations reveal the four-dimensional precipitation structure

Accurate four-dimensional (4D) precipitation information is essential for understanding the Earth's energy and water cycles, yet remains observationally unresolved at global scales. Conventional theory holds that geostationary infrared observations primarily sense cloud-top properties, with limited sensitivity to sub-cloud precipitation. Here we show that cloud-top infrared measurements nevertheless encode sufficient information to recover the four-dimensional structure of precipitation, revealing a previously unexploited observability of sub-cloud processes. We introduce a physically constrained deep learning framework, 4DPrecipNet, in which a moisture-first constraint requires the latent representation to recover precipitable water vapour, anchoring the model in thermodynamic consistency. By integrating multi-channel infrared radiances with these constraints and radar-derived precipitation profiles, we reconstruct the vertical and temporal evolution of precipitation systems from geostationary orbit. The framework captures deep convective structures and their evolution, with robust performance across large samples and independent radar comparisons. These results demonstrate that sub-cloud precipitation is physically encoded in cloud-top infrared observations, establishing a new pathway for continuous global monitoring of precipitation structure.

cs.CV↗

arXiv2026

423.7 + 426.5 Tb/s GMI Bi-Directional HCF Transmission

We demonstrate OESCL-band same-wavelength bi-directional transmission over 60 km HCF with 42.5 THz bandwidth, achieving GMIs comparable with the highest unidirectional SMF data-rates in both directions, with an aggregate of 423.7 + 426.5 Tb/s.

eess.SP↗

arXiv2026

Navigating Marginalization: Toward Justice-Oriented Socio-Technical Design for Parent-Child Learning among Southeast Asian Immigrant Mothers in Taiwan

This study investigates how Southeast Asian (SEA) immigrant mothers in Taiwan participate in their children's home-based learning. Drawing on semi-structured interviews and diary studies, we explore how these mothers navigate sociocultural constraints while fostering engagement and transmitting cultural values. Despite facing diminished agency and structural marginalization, mothers engage creatively in their children's everyday learning interactions. Guided by a justice-oriented lens, we identify various harms and propose design implications for socio-technical systems that center recognition, reciprocity, and accountability in parent-child learning at the individual, familial, and societal levels. Our contribution lies in foregrounding the role of intersectional identity in parent-child learning and proposing justice-oriented design directions that support the flourishing of immigrant mothers within socio-technical systems.

cs.HC↗

arXiv2026

"OpenBloom": A Question-Based LLM Tool to Support Stigma Reduction in Reproductive Well-Being

Reproductive well-being education remains widely stigmatized across diverse cultural contexts, constraining how individuals access and interpret reproductive health knowledge. We designed and evaluated OpenBloom, a stigma-sensitive, AI-mediated system that uses LLMs to transform reproductive health articles into reflective, question-based learning prompts. We employed OpenBloom as a design probe, aiming to explore the emerging challenges of reproductive well-being stigma through LLMs. Through surveys, semi-structured interviews, and focus group discussions, we examine how sociocultural stigma shapes participants' engagements with AI-generated questions and the opportunities of inquiry-based reproductive health education. Our findings identify key design considerations for stigma-sensitive LLM, including empathetic framing, inclusive language, values-based reflection, and explicit representation of marginalized identities. However, while current LLM outputs largely meet expectations for cultural sensitivity and non-offensiveness, they default to superficial rephrasing and factual recall rather than critical reflection. This guides well-being HCI design in sensitive health domains toward culturally grounded, participatory workflows.

cs.HC↗

arXiv2025

ERIENet: An Efficient RAW Image Enhancement Network under Low-Light Environment

RAW images have shown superior performance than sRGB images in many image processing tasks, especially for low-light image enhancement. However, most existing methods for RAW-based low-light enhancement usually sequentially process multi-scale information, which makes it difficult to achieve lightweight models and high processing speeds. Besides, they usually ignore the green channel superiority of RAW images, and fail to achieve better reconstruction performance with good use of green channel information. In this work, we propose an efficient RAW Image Enhancement Network (ERIENet), which parallelly processes multi-scale information with efficient convolution modules, and takes advantage of rich information in green channels to guide the reconstruction of images. Firstly, we introduce an efficient multi-scale fully-parallel architecture with a novel channel-aware residual dense block to extract feature maps, which reduces computational costs and achieves real-time processing speed. Secondly, we introduce a green channel guidance branch to exploit the rich information within the green channels of the input RAW image. It increases the quality of reconstruction results with few parameters and computations. Experiments on commonly used low-light image enhancement datasets show that ERIENet outperforms state-of-the-art methods in enhancing low-light RAW images with higher effiency. It also achieves an optimal speed of over 146 frame-per-second (FPS) for 4K-resolution images on a single NVIDIA GeForce RTX 3090 with 24G memory.

cs.CV↗

arXiv2025

Hyperflex: A SIMD-based DFA Model for Deep Packet Inspection

Deep Packet Inspection (DPI) has been extensively employed for network security. It examines traffic payloads by searching for regular expressions (regex) with the Deterministic Finite Automaton (DFA) model. However, as the network bandwidth and ruleset size are increasing rapidly, the conventional DFA model has emerged as a significant performance bottleneck of DPI. Leveraging the Single-Instruction-Multiple-Data (SIMD) instruction to perform state transitions can substantially boost the efficiency of the DFA model. In this paper, we propose Hyperflex, a novel SIMD-based DFA model designed for high-performance regex matching. Hyperflex incorporates a region detection algorithm to identify regions suitable for acceleration by SIMD instructions across the whole DFA graph. Also, we design a hybrid state transition algorithm that enables state transition in both SIMD-accelerated and normal regions, and ensures seamless state transition across the two types of regions. We have implemented Hyperflex on the commodity CPU and evaluated it with real network traffic and DPI regexes. Our evaluation results indicate that Hyperflex reaches a throughput of 8.89Gbit/s, representing an improvement of up to 2.27 times over Mcclellan, the default DFA model of the prominent multi-pattern regex matching engine Hyperscan. As a result, Hyperflex has been successfully deployed in Hyperscan, significantly enhancing its performance.

cs.NI↗

arXiv2025

AQUAH: Automatic Quantification and Unified Agent in Hydrology

We introduce AQUAH, the first end-to-end language-based agent designed specifically for hydrologic modeling. Starting from a simple natural-language prompt (e.g., 'simulate floods for the Little Bighorn basin from 2020 to 2022'), AQUAH autonomously retrieves the required terrain, forcing, and gauge data; configures a hydrologic model; runs the simulation; and generates a self-contained PDF report. The workflow is driven by vision-enabled large language models, which interpret maps and rasters on the fly and steer key decisions such as outlet selection, parameter initialization, and uncertainty commentary. Initial experiments across a range of U.S. basins show that AQUAH can complete cold-start simulations and produce analyst-ready documentation without manual intervention. The results are judged by hydrologists as clear, transparent, and physically plausible. While further calibration and validation are still needed for operational deployment, these early outcomes highlight the promise of LLM-centered, vision-grounded agents to streamline complex environmental modeling and lower the barrier between Earth observation data, physics-based tools, and decision makers.

cs.AI↗

arXiv2025

Gaze and Glow: Exploring Editing Processes on Social Media through Interactive Exhibition

We present Gaze and Glow, an interactive installation that reveals the often-invisible efforts of social media editing. Through narrative personas, experimental videos, and sensor-based interactions, the installation explores how audience attention shapes users' editing practices and emotional experiences. Deployed in a two-month public exhibition, Gaze and Glow engaged viewers and elicited responses. Reflexive thematic analysis of audience feedback highlights how making editing visible prompts new reflections on authenticity, agency, and performativity. We discuss implications for designing interactive systems that support selective memory, user-controlled visibility, and critical engagement with everyday digital self-presentation.

cs.HC↗

arXiv2025

Stable Polynomials via Undirected Colored Graphs

This paper initiates a systematic study of connections between undirected colored graphs and associated two-variable stable polynomials obtained via Cauchy transform-type formulas. Examples of such stable polynomials have played crucial roles in other recent studies, though their general properties have remained rather opaque. Using linear algebra techniques, this paper characterizes when these stable polynomials have boundary zeros and studies the finer behavior, called contact order, of their zero sets near the guaranteed boundary zeros.

math.CV↗

arXiv2025

Advances in Automated Fetal Brain MRI Segmentation and Biometry: Insights from the FeTA 2024 Challenge

Accurate fetal brain tissue segmentation and biometric analysis are essential for studying brain development in utero. The FeTA Challenge 2024 advanced automated fetal brain MRI analysis by introducing biometry prediction as a new task alongside tissue segmentation. For the first time, our diverse multi-centric test set included data from a new low-field (0.55T) MRI dataset. Evaluation metrics were also expanded to include the topology-specific Euler characteristic difference (ED). Sixteen teams submitted segmentation methods, most of which performed consistently across both high- and low-field scans. However, longitudinal trends indicate that segmentation accuracy may be reaching a plateau, with results now approaching inter-rater variability. The ED metric uncovered topological differences that were missed by conventional metrics, while the low-field dataset achieved the highest segmentation scores, highlighting the potential of affordable imaging systems when paired with high-quality reconstruction. Seven teams participated in the biometry task, but most methods failed to outperform a simple baseline that predicted measurements based solely on gestational age, underscoring the challenge of extracting reliable biometric estimates from image data alone. Domain shift analysis identified image quality as the most significant factor affecting model generalization, with super-resolution pipelines also playing a substantial role. Other factors, such as gestational age, pathology, and acquisition site, had smaller, though still measurable, effects. Overall, FeTA 2024 offers a comprehensive benchmark for multi-class segmentation and biometry estimation in fetal brain MRI, underscoring the need for data-centric approaches, improved topological evaluation, and greater dataset diversity to enable clinically robust and generalizable AI tools.

cs.CV↗

arXiv2025

On Robust Empirical Likelihood for Nonparametric Regression with Application to Regression Discontinuity Designs

Empirical likelihood serves as a powerful tool for constructing confidence intervals in nonparametric regression and regression discontinuity designs (RDD). The original empirical likelihood framework can be naturally extended to these settings using local linear smoothers, with Wilks' theorem holding only when an undersmoothed bandwidth is selected. However, the generalization of bias-corrected versions of empirical likelihood under more realistic conditions is non-trivial and has remained an open challenge in the literature. This paper provides a satisfactory solution by proposing a novel approach, referred to as robust empirical likelihood, designed for nonparametric regression and RDD. The core idea is to construct robust weights which simultaneously achieve bias correction and account for the additional variability introduced by the estimated bias, thereby enabling valid confidence interval construction without extra estimation steps involved. We demonstrate that the Wilks' phenomenon still holds under weaker conditions in nonparametric regression, sharp and fuzzy RDD settings. Extensive simulation studies confirm the effectiveness of our proposed approach, showing superior performance over existing methods in terms of coverage probabilities and interval lengths. Moreover, the proposed procedure exhibits robustness to bandwidth selection, making it a flexible and reliable tool for empirical analyses. The practical usefulness is further illustrated through applications to two real datasets.

math.ST↗

arXiv2025

On the Effectiveness of Large Language Models in Writing Alloy Formulas

Declarative specifications have a vital role to play in developing safe and dependable software systems. Writing specifications correctly, however, remains particularly challenging. This paper presents a controlled experiment on using large language models (LLMs) to write declarative formulas in the well-known language Alloy. Our use of LLMs is three-fold. One, we employ LLMs to write complete Alloy formulas from given natural language descriptions (in English). Two, we employ LLMs to create alternative but equivalent formulas in Alloy with respect to given Alloy formulas. Three, we employ LLMs to complete sketches of Alloy formulas and populate the holes in the sketches by synthesizing Alloy expressions and operators so that the completed formulas accurately represent the desired properties (that are given in natural language). We conduct the experimental evaluation using 11 well-studied subject specifications and employ two popular LLMs, namely ChatGPT and DeepSeek. The experimental results show that the LLMs generally perform well in synthesizing complete Alloy formulas from input properties given in natural language or in Alloy, and are able to enumerate multiple unique solutions. Moreover, the LLMs are also successful at completing given sketches of Alloy formulas with respect to natural language descriptions of desired properties (without requiring test cases). We believe LLMs offer a very exciting advance in our ability to write specifications, and can help make specifications take a pivotal role in software development and enhance our ability to build robust software.

cs.SE↗

arXiv2025

Balancing Sleep and Study: Cultural Contexts in Family Informatics for Taiwanese Parents and Children

This study examines the intersection of academic pressure and sleep within Taiwanese families, revealing how cultural norms and expectations shape sleep practices. Through interviews and two-week diaries from eleven families, we found that academic demands significantly influence children's sleep patterns, leading to reduced sleep duration and varied sleep schedules. Our research highlights the importance of integrating care and attuning into the design of sleep-tracking technologies, advocating for a family informatics approach that considers both health needs and social expectations. By exploring these dynamics, we contribute to a broader understanding of family contexts in diverse cultural settings and offer insights for more inclusive technology design.

cs.HC↗