arXiv subjects

Xinyue Chen

Publications and source records attributed to Xinyue Chen.

At least 19 recordsLinked to original sources

arXiv2026

Layer-Hybridized Wigner Crystals in MoSe2/WS2 Moir\'e Superlattice

Transition metal dichalcogenide moir\'e heterobilayers with type-II band alignment provide a versatile platform for layer-polarized generalized Wigner crystals, in which strong Coulomb interactions drive charge ordering at fractional lattice fillings. With a finite interlayer band offset, an out-of-plane electric field can tune layer-resolved moir\'e bands through resonance and enable controllable interlayer hybridization. Although hybridized Mott insulators have been previously demonstrated, whether fractional charge-ordered states can survive such hybridization remains elusive. Here we drive an H-stacked MoSe2/WS2 moir\'e heterobilayer through a type-I-to-type-II band-alignment transition and realize layer-hybridized Mott insulator and generalized Wigner crystals. For fillings below one electron per moir\'e cell, tunneling delocalizes electrons and modifies Wigner crystallization. However, above one electron per cell, Coulomb repulsion overcomes tunneling and favors layer-separated occupation, stabilizing stronger charge-ordered states. These results establish electrically tunable hybridized moir\'e heterobilayers as a powerful platform for engineering correlated charge order and exploring fractional Chern phases and emergent magnetism.

cond-mat.mes-hall↗

arXiv2026

ICAConfPubs: A Dataset and User Interface for ICA Conference Papers (2003-2018)

This paper presents a comprehensive dataset of past ICA (The International Communication Association) annual conference papers from 2003 to 2018, encompassing 27,466 papers, 21,038 authors, and 4,935 sessions. We made the dataset publicly available in both CSV and JSON formats. Additionally, we developed an API to facilitate programmatic access, and an intuitive user interface to enable users to navigate and explore the data more easily. The web application, API documentation, downloadable data, and reproducible code to obtain and process the data are available at https://ica.hongtaoh.com.

cs.DL↗

arXiv2026

Yuvion LLM: An Adversarially-Aware Large Language Model for Content And AI Safety

As large language models are increasingly deployed in real-world systems, safety failures can still lead to harmful outputs and dangerous misuse. We argue that the essence of safety is adversarial: many failures arise not from natural inputs alone, but from strategic attempts to evade model policies and safeguards. However, existing general-purpose model development largely overlook this adversarial nature, and often remain insufficient for realistic safety scenarios involving planning, tool use, and multi-step reasoning, causing measured safety performance to overestimate real deployment robustness. To address this gap, we present Yuvion LLM, a large language model built for adversarially robust content safety and broader AI safety. Yuvion LLM treats adversarial robustness and agentic capability as first-class objectives. Its pipeline combines adversarially aware data construction, knowledge-enhanced continued pretraining, and policy-grounded multi-task safety post-training, including risk-aware supervised fine-tuning and reinforcement learning-based policy optimization, together with safety-aware agentic reinforcement learning for tool use and multi-step reasoning in complex safety scenarios. We further introduce the Yuvion LLM RiskEval (YLRE), a collection of 93 benchmarks across four evaluation categories, covering diverse open and internal evaluations with a focus on safety, adversarial robustness, and real-world capability requirements. Across these evaluations, Yuvion LLM demonstrates clear advantages on safety-focused benchmarks and particularly strong robustness under adversarial conditions, while maintaining solid overall capability. Notably, Yuvion-8B outperforms most state-of-the-art baselines, including substantially larger models such as GPT-5.4 and Qwen3-MAX, on several safety tasks.

cs.CL↗

arXiv2026

Yuvion VL: A Multimodal Foundation Model for Adversarial Content and AI Safety

General-purpose models often struggle to reliably identify and understand real-world multimodal risks, largely due to the inherent multimodal adversarial nature of content and AI safety. We present Yuvion VL, a family of multimodal large language models purpose-built for content and AI safety, with both instruction-tuned and reasoning-oriented variants. Yuvion VL addresses this gap by treating safety as an inherently adversarial and multimodal problem and designing the entire pipeline around adversarial robustness. For data construction, we develop an automated pipeline integrating adversarial-aware data synthesis with multi-stage quality control, producing large-scale, high-quality multimodal samples augmented with domain knowledge and reasoning annotations. For training, we adopt a three-stage pipeline that includes continued pretraining for risk-concept cross-modal alignment, instruct post-training for production-grade safety tasks, and reasoning post-training for enhanced interpretability and performance in complex tasks. We further introduce Confuse-then-Contrast Fine-Tuning, a contrastive framework that mines model-specific confusions and constructs multi-image contrastive groups to enforce explicit discrimination of fine-grained visual-semantic elements, enabling the model to distinguish between visually similar cases with different safety implications in adversarial safety tasks. To support rigorous evaluation, we further introduce Yuvion VL RiskEval (YVRE), a collection of benchmarks covering diverse open and internal evaluations, with a focus on content and AI safety, adversarial robustness, and real-world capability requirements. Experiments show that Yuvion VL-32B achieves industry-leading safety performance, surpassing comparably sized open-source models and best closed-source commercial models, while maintaining comparable general capabilities.

cs.CV↗

arXiv2026

CoAction: Cross-task Correlation-aware Pareto Set Learning

Pareto set learning (PSL) is an emerging paradigm in multi-objective optimization that trains neural networks to map preference vectors to Pareto optimal solutions. However, existing PSL methods primarily focus on solving a single multi-objective optimization problem at a time. This limitation not only increases computational costs in multi-objective multitask optimization scenarios by requiring a separate model for each task, but also fails to exploit the inter-task correlations across tasks. To address this, we propose a Cross-tAsk correlation-aware Pareto Set Learning (CoAction) framework, which leverages task-aware transformer to handle multiple tasks simultaneously. Specifically, by assigning task-specific embedding vectors to individual tasks, the model effectively distinguishes between tasks while facilitating knowledge sharing among them. We utilize a Transformer encoder as the backbone architecture to leverage its self-attention mechanism for capturing complex task dependencies. The proposed approach is evaluated on comprehensive multitask test suites covering both benchmark problems and real-world applications, demonstrating effectiveness and competitive performance in Hypervolume, Range, and Sparsity.

cs.LG↗

arXiv2026

How Historians Use Visualization: A Corpus-Driven Mixed-Methods Study

Visualization in historical research is shifting from isolated attempts to systematic practices. However, data-driven evidence about how historians actually use visualization remains scarce. We present a corpus-driven, mixed-methods study that combines analysis of images from 4,142 research articles across history and digital humanities journals with a collaboratively developed visualization taxonomy and a semi-automatic labeling pipeline. We construct a corpus of 14,021 images, classify 4,831 visualization instances using a hierarchical, domain-informed taxonomy, and analyze patterns of visualization adoption across venues, history subfields, and time. To interpret these patterns, we conduct interviews with 11 historians and use HiFigAtlas system as a boundary object to support joint inspection of the corpus. We identify distinct roles for visualizations in historical research: primary-source, evidence-synthesis, communicative, confirmative, and exploratory. We further find that while historians pursue diverse goals with figures, persistent epistemological and practical barriers, such as uncertainty, provenance, justification burden, and publication constraints, impede the adoption of visualization. This work contributes a grounded account of visualization use in historical scholarship and points to opportunities to better support domain-specific needs.

cs.GR↗

arXiv2026

Can LLM-Generated Text Empower Surgical Vision-Language Pre-training?

Recent advancements in self-supervised learning have led to powerful surgical vision encoders capable of spatiotemporal understanding. However, extending these visual foundations to multi-modal reasoning tasks is severely bottlenecked by the prohibitive cost of expert textual annotations. To overcome this scalability limitation, we introduce \textbf{LIME}, a large-scale multi-modal dataset derived from open-access surgical videos using human-free, Large Language Model (LLM)-generated narratives. While LIME offers immense scalability, unverified generated texts may contain errors, including hallucinations, that could potentially lead to catastrophically degraded pre-trained medical priors in standard contrastive pipelines. To mitigate this, we propose \textbf{SurgLIME}, a parameter-efficient Vision-Language Pre-training (VLP) framework designed to learn reliable cross-modal alignments using noisy narratives. SurgLIME preserves foundational medical priors using a LoRA-adapted dual-encoder architecture and introduces an automated confidence estimation mechanism that dynamically down-weights uncertain text during contrastive alignment. Evaluations on the AutoLaparo and Cholec80 benchmarks show that SurgLIME achieves competitive zero-shot cross-modal alignment while preserving the robust linear probing performance of the visual foundation model. Dataset, code, and models are publicly available at https://github.com/visurg-ai/SurgLIME.

cs.CV↗

arXiv2026

ConSearcher: Supporting Conversational Information Seeking in Online Communities with Member Personas

Many people browse online communities to learn from others' experiences and opinions, e.g., for constructing travel plans. Conversational search powered by large language models (LLMs) could ease this information-seeking task, but it remains under-investigated within the online community. In this paper, we first conducted an exploratory study (N=10) that indicated the helpfulness of a classic conversational search tool and identified room for improvement. Then, we proposed ConSearcher, an LLM-powered tool with dynamically generated member personas based on user queries to facilitate conversational search in the community. In ConSearcher, users can clarify their interests by checking what a simulated member similar to them may ask and get responses from diverse members' perspectives. A within-subjects study (N=27) showed that compared to two conversational search baselines, ConSearcher led to significantly higher information-seeking outcome and user engagement but raised concerns about over-personalization. We discuss implications for supporting conversational information seeking in online communities.

cs.HC↗

arXiv2026

MIU2Net: weak-lensing mass inversion using deep learning with nested U-structures

One of the primary goals of next-generation gravitational lensing surveys is to measure the large-scale distribution of dark matter, which requires accurate mass inversion to convert weak-lensing shear maps into convergence (kappa) fields. This work develops a mass inversion method tailored for upcoming space missions such as CSST and Euclid, aiming to recover both the mass distribution and the convergence power spectrum with high fidelity. We introduce MIU2Net, a versatile deep-learning framework for kappa-map reconstruction based on the U2-Net architecture. A new loss function is constructed to jointly estimate the convergence field and its frequency-domain energy distribution, effectively balancing optimal mean squared error and optimal power-spectrum recovery. The method incorporates realistic observational effects into shear fields, including shape noise, reduced shear, and complex masks. Under noise levels anticipated for future space-based lensing surveys, MIU2Net recovers the convergence power spectrum with 4% uncertainties up to l approximately 500, significantly outperforming Wiener filtering and MCALens. Beyond two-point statistics, the method accurately reconstructs the convergence distribution, peak centroid, and peak amplitude. Compared to other learning-based approaches such as DeepMass, MIU2Net reduces the root-mean-square error by 5% without smoothing and by 38% with a 1-arcmin smoothing scale. MIU2Net represents a substantial advancement in mass inversion methodology, offering improved accuracy in both RMSE and power-spectrum reconstruction. It provides a promising tool for mapping dark matter environments and large-scale structures in the era of next-generation space lensing surveys.

astro-ph.CO↗

arXiv2025

Back to the Feature: Explaining Video Classifiers with Video Counterfactual Explanations

Counterfactual explanations (CFEs) are minimal and semantically meaningful modifications of the input of a model that alter the model predictions. They highlight the decisive features the model relies on, providing contrastive interpretations for classifiers. State-of-the-art visual counterfactual explanation methods have primarily focused on interpreting image classifiers, leaving the domain of video models relatively underexplored. For the video CFEs to be useful, they have to be physically plausible, temporally coherent, and exhibit smooth motion trajectories. Existing CFE image-based methods, designed to explain image classifiers, lack the capacity to generate temporally coherent, smooth and physically plausible video CFEs. To address this, we propose Back To The Feature (BTTF), an optimization framework that generates video CFEs. Our method introduces two novel features, 1) an optimization scheme to retrieve the initial latent noise conditioned by the first frame of the input video, 2) a two-stage optimization strategy to enable the search for counterfactual videos in the vicinity of the input video. Both optimization processes are guided solely by the target classifier, ensuring the explanation is faithful. To accelerate convergence, we also introduce a progressive optimization strategy that incrementally increases the number of denoising steps. Extensive experiments on video datasets such as Shape-Moving (motion classification), MEAD (emotion classification), and NTU RGB+D (action classification) show that our BTTF effectively generates valid, visually similar and realistic counterfactual videos that provide concrete insights into the classifier's decision-making mechanism.

cs.CV↗

arXiv2025

A Stitch in Time: Learning Procedural Workflow via Self-Supervised Plackett-Luce Ranking

Procedural activities, ranging from routine cooking to complex surgical operations, are highly structured sequences of actions performed in a specific temporal order. Despite the success of current self-supervised learning (SSL) methods on static images and short clips, these models often overlook the underlying sequential structure of such activities. We expose this lack of procedural awareness with a motivating experiment: models pretrained on forward and time-reversed sequences produce highly similar features, confirming that their representations are blind to the underlying procedural order. To address this shortcoming, we propose PL-Stitch, a self-supervised framework that harnesses the inherent temporal order of video frames as a powerful supervisory signal. Our approach integrates two novel probabilistic objectives based on the Plackett-Luce (PL) model. The primary PL objective trains the model to sort sampled frames chronologically, compelling it to learn the global workflow progression. The secondary objective, a spatio-temporal jigsaw loss, complements the learning by capturing fine-grained, cross-frame object correspondences. Our approach consistently achieves superior performance across five surgical and cooking benchmarks. Specifically, PL-Stitch yields significant gains in surgical phase recognition (e.g., +11.4 pp in k-NN accuracy on Cholec80) and cooking action segmentation (e.g., +5.7 pp in linear probing accuracy on Breakfast), demonstrating its effectiveness for procedural video representation learning. Code and models are available at https://github.com/visurg-ai/PL-Stitch.

cs.CV↗

arXiv2025

A Proof of the 2004 Albert-Grossman-Nowakowski-Wolfe Conjecture on Alternating Linear Clobber

Clobber is an alternate-turn two-player game introduced in 2001 by Albert, Grossman, Nowakowski and Wolfe. The board is a graph with each node colored black (x), white (o), or empty (-). Player Left has black stones, player Right has white stones. On a turn, a player takes one of their stones that is adjacent to an opponent stone and clobbers the opponent's stone (replaces it with theirs). Whoever cannot move loses. Linear clobber is clobber played on a path, for example, one row of a Go board. In 2004 Albert et al. conjectured that, for every even-length alternating-color linear clobber position except oxoxox, the first player has a winning strategy. We prove their conjecture.

math.CO↗

arXiv2025

More AI Assistance Reduces Cognitive Engagement: Examining the AI Assistance Dilemma in AI-Supported Note-Taking

As AI tools become increasingly embedded in cognitively demanding tasks such as note-taking, questions remain about whether they enhance or undermine cognitive engagement. This paper examines the "AI Assistance Dilemma" in note-taking, investigating how varying levels of AI support affect user engagement and comprehension. In a within-subject experiment, we asked participants (N=30) to take notes during lecture videos under three conditions: Automated AI (high assistance with structured notes), Intermediate AI (moderate assistance with real-time summary, and Minimal AI (low assistance with transcript). Results reveal that Intermediate AI yields the highest post-test scores and Automated AI the lowest. Participants, however, preferred the automated setup due to its perceived ease of use and lower cognitive effort, suggesting a discrepancy between preferred convenience and cognitive benefits. Our study provides insights into designing AI assistance that preserves cognitive engagement, offering implications for designing moderate AI support in cognitive tasks.

cs.HC↗

arXiv2025

A Comparative Study of Halo Mass Estimates from Group Catalogs and Lensing Signals

We compare halo mass estimates from three galaxy group catalogs (redMaPPer, Yang21, and Zou21) with those derived from gravitational lensing measurements. Each catalog employs distinct methodologies, including mass-richness relations, abundance matching, and luminosity-based calibration. A linear correlation is observed between catalog-estimated and lensing-derived masses. The redMaPPer catalog shows the best agreement, especially for lower-redshift groups, with minor deviations in higher-redshift bins. Yang21 is the only catalog containing low mass groups, which gives a reasonably good mass estimation, except for the lowest mass bin. Cross-matched groups between redMaPPer and Yang21 reveal the former catalog provides more accurate mass estimation, while the Yang21 makes under-estimation of halo mass for those sharing the central galaxy with redMaPPer and over-estimation of halo mass for those with different center determination with redMaPPer and for the unique Yang21 groups. These findings emphasize the importance of redshift-dependent calibration and refined group definitions for accurate mass estimation.

astro-ph.CO↗

arXiv2025

TAMMs: Change Understanding and Forecasting in Satellite Image Time Series with Temporal-Aware Multimodal Models

Temporal Change Description (TCD) and Future Satellite Image Forecasting (FSIF) are critical, yet historically disjointed tasks in Satellite Image Time Series (SITS) analysis. Both are fundamentally limited by the common challenge of modeling long-range temporal dynamics. To explore how to improve the performance of methods on both tasks simultaneously by enhancing long-range temporal understanding capabilities, we introduce **TAMMs**, the first unified framework designed to jointly perform TCD and FSIF within a single MLLM-diffusion architecture. TAMMs introduces two key innovations: Temporal Adaptation Modules (**TAM**) enhance frozen MLLM's ability to comprehend long-range dynamics, and Semantic-Fused Control Injection (**SFCI**) mechanism translates this change understanding into fine-grained generative control. This synergistic design makes the understanding from the TCD task to directly inform and improve the consistency of the FSIF task. Extensive experiments demonstrate TAMMs significantly outperforms state-of-the-art specialist baselines on both tasks. Our dataset can be found at https://huggingface.co/datasets/IceInPot/TAMMs .

cs.CV↗

arXiv2025

What Does Success Look Like? Catalyzing Meeting Intentionality with AI-Assisted Prospective Reflection

Despite decades of HCI and Meeting Science research, complaints about ineffective meetings are still pervasive. We argue that meeting technologies lack support for prospective reflection, that is, thinking about why a meeting is needed and what might happen. To explore this, we designed a Meeting Purpose Assistant (MPA) technology probe to coach users to articulate their meeting's purpose and challenges, and act accordingly. The MPA used Generative AI to support personalized and actionable prospective reflection across the diversity of meeting contexts. Using a participatory prompting methodology, 18 employees of a global technology company reflected with the MPA on upcoming meetings. Observed impacts were: clarifying meeting purposes, challenges, and success conditions; changing perspectives and flexibility; improving preparation and communication; and proposing changed plans. We also identify perceived social, temporal, and technological barriers to using the MPA. We present system and workflow design considerations for developing AI-assisted reflection support for meetings.

cs.HC↗

arXiv2025

VisTaxa: Developing a Taxonomy of Historical Visualizations

Historical visualizations are a rich resource for visualization research. While taxonomy is commonly used to structure and understand the design space of visualizations, existing taxonomies primarily focus on contemporary visualizations and largely overlook historical visualizations. To address this gap, we describe an empirical method for taxonomy development. We introduce a coding protocol and the VisTaxa system for taxonomy labeling and comparison. We demonstrate using our method to develop a historical visualization taxonomy by coding 400 images of historical visualizations. We analyze the coding result and reflect on the coding process. Our work is an initial step toward a systematic investigation of the design space of historical visualizations.

cs.HC↗

arXiv2025

Redesign of Online Design Communities: Facilitating Personalized Visual Design Learning with Structured Comments

Online Design Communities (ODCs) offer various artworks with members' comments for beginners to learn visual design. However, as identified by our Formative Study (N = 10), current ODCs lack features customized for personal learning purposes, e.g., searching artworks and digesting useful comments to learn design principles about buttons. In this paper, we present DesignLearner, a redesigned interface of ODCs to facilitate personalized visual design learning with comments structured based on UI components (e.g., button, text) and visual elements (e.g., color, contrast). In DesignLearner, learners can specify the UI components and visual elements that they wish to learn to filter artworks and associated comments. They can interactively read comments on an artwork, take notes, and get suggestions for the next artworks to explore. Our between-subjects study (N = 24) indicates that compared to a traditional ODC interface, DesignLearner can improve the user learning outcome and is deemed significantly more useful. We conclude with design considerations for customizing the interface of online communities to satisfy users' learning needs.

cs.HC↗