arXiv subjects

Yutong Yang

Publications and source records attributed to Yutong Yang.

At least 19 recordsLinked to original sources

arXiv2026

Imagine Before Retrieval: Prospective Skill Retrieval for LLM Agents

Skill retrieval has recently emerged as a promising paradigm for identifying the desirable execution guidelines from the skill gallery, thus equipping large language model (LLM) agents with the procedural knowledge to accomplish the specified task. To this end, most existing methods customize the retrieval model or reconfigure the retrieval pipeline to prioritize skills that are most semantically relevant to the task query. However, we empirically reveal that task queries and skills are naturally formulated from different perspectives, namely, objective-oriented and procedural-oriented, leading to an under-explored problem termed Query--Skill Misalignment (QSM). Clearly, it is daunting and even impossible to associate the desirable skills in the context of QSM, thus hindering the agent from correctly executing the task. As a remedy, inspired by human prospective cognition, we propose SkillDreamer, a novel framework to alleviate the negative impact of QSM problem. In brief, SkillDreamer first infers the capabilities necessary for task execution, then imagines how to realize these capabilities by generating pseudo skills, and finally leverages such prospective information to bridge the gap between objective-oriented task queries and execution-oriented skills. Extensive experiments on SkillRet and SkillUsage not only verify the effectiveness of SkillDreamer in both skill retrieval and end-to-end task execution, but also demonstrate its generalizability across diverse retrieval models and pipelines. The code will be released upon acceptance.

cs.IR↗

arXiv2026

PixelControl: Fine-Grained Condition Fidelity in Text-to-Image Diffusion

Controllable text-to-image diffusion models can often follow the global layout of spatial conditions, yet still violate fine-grained structures such as object boundaries, thin contours, and medium/small conditioned regions. This limitation is especially problematic for VAE-based latent diffusion, where spatial compression can weaken high-frequency and low-area condition signals. We propose PixelControl, a pixel-space controllable diffusion framework for fine-grained condition fidelity. Built on a PixelDiT-style backbone, PixelControl avoids the latent bottleneck and introduces two complementary designs. First, Structure-Aware Control Injection derives a condition structure map and uses it to strengthen injected control residuals around spatially sensitive regions. Second, Multi-Scale Pyramid Cycle Loss verifies generated images against condition-derived structures across multiple resolutions, balancing global layout consistency with local boundary and detail accuracy. PixelControl supports depth, segmentation, edge, and their combinations through modality-specific control branches with lightweight gated fusion. Experiments across depth, segmentation, and edge control show that PixelControl improves structural fidelity and visual quality over existing controllable generation methods, with especially strong gains on boundaries and medium/small conditioned regions. The project page can be found at: https://linxin0.github.io/pixelcontrol_homepage/pixelcontrol-site/

cs.CV↗

arXiv2026

Beyond Symmetric Fusion: Exploiting Task-Dependent Modality Strengths for RGB-Event Small Object Detection

State-of-the-art RGB-Event detectors improve the detection of small, fast-moving objects by combining complementary features from RGB and Event data, yet they typically fuse the two modalities into a unified representation for both localization and classification. Such a task-symmetric design is inconsistent with the intuition that the two modalities should play different roles according to their task-specific strengths. To examine this issue, we conduct a modality-specific evaluation and find that the relative advantage of the two modalities reverses across tasks: Event data are substantially more effective for class-agnostic localization, whereas RGB data provide stronger category evidence within localized target regions. Motivated by this task-dependent asymmetry, we propose an Asymmetric Event-RGB Object Detection Transformer (AERODet). During class-agnostic localization, Scale-wise Uncertainty-aware Reliability Estimation (SURE) calculates the relative reliability of the two modalities from their objectness response heatmaps and accordingly calibrates their contributions when the decoder aggregates multimodal features. Once the candidate boxes are obtained, Task-Decoupled Semantic Refinement (TDSR) decouples classification from localization and uses RGB RoI features for fine-grained classification. Extensive experiments on FRED and NeRDD demonstrate that AERODet achieves state-of-the-art performance. In particular, it surpasses the strongest RGB-Event baseline by 10.7 mAP points on the FRED challenging split.

cs.CV↗

arXiv2026

ActTraitBench: Quantifying the Knowledge-Decision Gap in Large Language Models via Human-Grounded Behavioral Validation

While Large Language Models (LLMs) can convincingly simulate personas in explicit self-reports, they often deviate in implicit behavioral decisions, revealing a substantial Knowledge-Decision Gap ($G_{\mathrm{KD}}$). Existing benchmarks struggle to measure this discrepancy due to limited construct validity, multidimensional entanglement, and distributional biases in LLM-based evaluation. To address these issues, we propose ActTraitBench, a human-grounded evaluation framework for measuring personality consistency in LLMs. Grounded in empirical human data, ActTraitBench establishes one-to-one mappings between psychometric facets and behavioral paradigms and applies Distributional Calibration via Quantile Mapping to reduce distributional mismatch between LLM-judge scores and human responses. Experiments on 14 mainstream LLMs reveal substantial knowledge-decision gaps and show that assigned personas are reflected more consistently in self-reports than in behavioral decisions for most evaluated models. To mitigate this gap, we further introduce the Chain of Cognitive Alignment (CoCA), an inference-time intervention that reduces $G_{\mathrm{KD}}$ for 12 of the 13 models with paired results. Code and resources are available at https://github.com/Selina233/ActTraitBench.

cs.CL↗

arXiv2026

Aligning Language Models with Real-time Knowledge Editing

Knowledge editing aims to modify outdated knowledge in language models efficiently while retaining their original capabilities. Mainstream datasets for knowledge editing are predominantly static and fail to keep in pace with the evolving real-world knowledge. In this work, we introduce CRAFT, an ever-evolving real-world dataset for knowledge editing. It evaluates models on temporal locality, common-sense locality, composite portability and alias portability, providing a comprehensive and challenging evaluation for knowledge editing, on which previous methods hardly achieve balanced performance. Towards flexible real-time knowledge editing, we propose KEDAS, a novel paradigm of knowledge editing alignment featuring diverse edit augmentation and self-adaptive post-alignment inference, exhibiting significant performance gain on both CRAFT and traditional datasets compared to previous methods. We hope this work may serve as a catalyst for shifting the focus of knowledge editing from static update to dynamic evolution.

cs.CL↗

arXiv2026

Individual and Combined Effects of English as a Second Language and Typos on LLM Performance

Large language models (LLMs) are used globally, and because much of their training data is in English, they typically perform best on English inputs. As a result, many non-native English speakers interact with them in English as a second language (ESL), and these inputs often contain typographical errors. Prior work has largely studied the effects of ESL variation and typographical errors separately, even though they often co-occur in real-world use. In this study, we use the Trans-EnV framework to transform standard English inputs into eight ESL variants and apply MulTypo to inject typos at three levels: low, moderate, and severe. We find that combining ESL variation and typos generally leads to larger performance drops than either factor alone, though the combined effect is not simply additive. This pattern is clearest on closed-ended tasks, where performance degradation can be characterized more consistently across ESL variants and typo levels, while results on open-ended tasks are more mixed. Overall, these findings suggest that evaluations on clean standard English may overestimate real-world model performance, and that evaluating ESL variation and typographical errors in isolation does not fully capture model behavior in realistic settings.

cs.CL↗

arXiv2026

Do Emotions in Prompts Matter? Effects of Emotional Framing on Large Language Models

Emotional tone is pervasive in human communication, yet its influence on large language model (LLM) behaviour remains unclear. Here, we examine how first-person emotional framing in user-side queries affect LLM performance across six benchmark domains, including mathematical reasoning, medical question answering, reading comprehension, commonsense reasoning and social inference. Across models and tasks, static emotional prefixes usually produce only small changes in accuracy, suggesting that affective phrasing is typically a mild perturbation rather than a reliable general-purpose intervention. This stability is not uniform: effects are more variable in socially grounded tasks, where emotional context more plausibly interacts with interpersonal reasoning. Additional analyses show that stronger emotional wording induces only modest extra change, and that human-written prefixes reproduce the same qualitative pattern as LLM-generated ones. We then introduce EmotionRL, an adaptive emotional prompting framework that selects emotional framing adaptively for each query. Although no single emotion is consistently beneficial, adaptive selection yields more reliable gains than fixed emotional prompting. Together, these findings show that emotional tone is neither a dominant driver of LLM performance nor irrelevant noise, but a weak and input-dependent signal that can be exploited through adaptive control.

cs.AI↗

arXiv2026

CFMS: Towards Explainable and Fine-Grained Chinese Multimodal Sarcasm Detection Benchmark

Multimodal sarcasm detection has recently garnered significant attention. However, existing benchmarks suffer from coarse-grained annotations and limited cultural coverage, which hinder research into fine-grained semantic understanding. To address this, we construct CFMS, the first fine-grained multimodal sarcasm dataset tailored for Chinese social media. It comprises 2,796 high-quality image-text pairs and provides a triple-level annotation framework: sarcasm identification, target recognition, and explanation generation. We find that the fine-grained explanation annotations effectively guide AI in generating images with explicit sarcastic intent. Furthermore, we curate a high-consistency parallel Chinese-English metaphor subset (200 entries each), revealing significant limitations of current models in metaphoric reasoning. To overcome the constraints of traditional retrieval methods, we propose a Reinforcement Learning-augmented In-Context Learning strategy (PGDS) to dynamically optimize exemplar selection. Extensive experiments demonstrate that CFMS provides a solid foundation for building reliable multimodal sarcasm understanding systems, and the PGDS method significantly outperforms existing baselines on key tasks. Our data and code are available at https://anonymous.4open.science/r/CFMS-E8F9.

cs.CL↗

arXiv2026

GroupEnsemble: Efficient Uncertainty Estimation for DETR-based Object Detection

Detection Transformer (DETR) and its variants show strong performance on object detection, a key task for autonomous systems. However, a critical limitation of these models is that their confidence scores only reflect semantic uncertainty, failing to capture the equally important spatial uncertainty. This results in an incomplete assessment of the detection reliability. On the other hand, Deep Ensembles can tackle this by providing high-quality spatial uncertainty estimates. However, their immense memory consumption makes them impractical for real-world applications. A cheaper alternative, Monte Carlo (MC) Dropout, suffers from high latency due to the need of multiple forward passes during inference to estimate uncertainty. To address these limitations, we introduce GroupEnsemble, an efficient and effective uncertainty estimation method for DETR-like models. GroupEnsemble simultaneously predicts multiple individual detection sets by feeding additional diverse groups of object queries to the transformer decoder during inference. Each query group is transformed by the shared decoder in isolation and predicts a complete detection set for the same input. An attention mask is applied to the decoder to prevent inter-group query interactions, ensuring each group detects independently to achieve reliable ensemble-based uncertainty estimation. By leveraging the decoder's inherent parallelism, GroupEnsemble efficiently estimates uncertainty in a single forward pass without sequential repetition. We validated our method under autonomous driving scenes and common daily scenes using the Cityscapes and COCO datasets, respectively. The results show that a hybrid approach combining MC-Dropout and GroupEnsemble outperforms Deep Ensembles on several metrics at a fraction of the cost. The code is available at https://github.com/yutongy98/GroupEnsemble.

cs.CV↗

arXiv2026

Single-Slice-to-3D Reconstruction in Medical Imaging and Natural Objects: A Comparative Benchmark with SAM 3D

While three-dimensional imaging is essential for clinical diagnosis, its high cost and long wait times have motivated the use of image-to-3D foundation models to infer volume from two-dimensional modalities. However, because these models are trained on natural images, their learned geometric priors struggle to transfer to inherently planar medical data. A benchmark of five state-of-the-art models (SAM3D, Hunyuan3D-2.1, Direct3D, Hi3DGen, and TripoSG) across six medical and two natural datasets revealed that voxel-based overlap remains uniformly low across all methods due to severe depth ambiguity from single-slice inputs. Despite this fundamental volumetric failure, global distance metrics indicate that SAM3D best captures topological similarity to ground-truth medical shapes, whereas alternative models are prone to oversimplification. Ultimately, these findings quantify the limits of zero-shot single-slice 3D inference, highlighting that reliable medical 3D reconstruction requires domain-specific adaptation and anatomical constraints to overcome complex medical geometries.

cs.CV↗

arXiv2026

Reliable Thinking with Images

As a multimodal extension of Chain-of-Thought (CoT), Thinking with Images (TWI) has recently emerged as a promising avenue to enhance the reasoning capability of Multi-modal Large Language Models (MLLMs), which generates interleaved CoT by incorporating visual cues into the textual reasoning process. However, the success of existing TWI methods heavily relies on the assumption that interleaved image-text CoTs are faultless, which is easily violated in real-world scenarios due to the complexity of multimodal understanding. In this paper, we reveal and study a highly-practical yet under-explored problem in TWI, termed Noisy Thinking (NT). Specifically, NT refers to the imperfect visual cues mining and answer reasoning process. As the saying goes, ``One mistake leads to another'', erroneous interleaved CoT would cause error accumulation, thus significantly degrading the performance of MLLMs. To solve the NT problem, we propose a novel method dubbed Reliable Thinking with Images (RTWI). In brief, RTWI estimates the reliability of visual cues and textual CoT in a unified text-centric manner and accordingly employs robust filtering and voting modules to prevent NT from contaminating the final answer. Extensive experiments on seven benchmarks verify the effectiveness of RTWI against NT.

cs.CV↗

arXiv2026

Grading Scale Impact on LLM-as-a-Judge: Human-LLM Alignment Is Highest on 0-5 Grading Scale

Large language models (LLMs) are increasingly used as automated evaluators, yet prior works demonstrate that these LLM judges often lack consistency in scoring when the prompt is altered. However, the effect of the grading scale itself remains underexplored. We study the LLM-as-a-judge problem by comparing two kinds of raters: humans and LLMs. We collect ratings from both groups on three scales and across six benchmarks that include objective, open-ended subjective, and mixed tasks. Using intraclass correlation coefficients (ICC) to measure absolute agreement, we find that LLM judgments are not perfectly consistent across scales on subjective benchmarks, and that the choice of scale substantially shifts human-LLM agreement, even when within-group panel reliability is high. Aggregated over tasks, the grading scale of 0-5 yields the strongest human-LLM alignment. We further demonstrate that pooled reliability can mask benchmark heterogeneity and reveal systematic subgroup differences in alignment across gender groups, strengthening the importance of scale design and sub-level diagnostics as essential components of LLM-as-a-judge protocols.

cs.CL↗

arXiv2025

SpaceDrive: Infusing Spatial Awareness into VLM-based Autonomous Driving

End-to-end autonomous driving methods built on vision language models (VLMs) have undergone rapid development driven by their universal visual understanding and strong reasoning capabilities obtained from the large-scale pretraining. However, we find that current VLMs struggle to understand fine-grained 3D spatial relationships which is a fundamental requirement for systems interacting with the physical world. To address this issue, we propose SpaceDrive, a spatial-aware VLM-based driving framework that treats spatial information as explicit positional encodings (PEs) instead of textual digit tokens, enabling joint reasoning over semantic and spatial representations. SpaceDrive employs a universal positional encoder to all 3D coordinates derived from multi-view depth estimation, historical ego-states, and text prompts. These 3D PEs are first superimposed to augment the corresponding 2D visual tokens. Meanwhile, they serve as a task-agnostic coordinate representation, replacing the digit-wise numerical tokens as both inputs and outputs for the VLM. This mechanism enables the model to better index specific visual semantics in spatial reasoning and directly regress trajectory coordinates rather than generating digit-by-digit, thereby enhancing planning accuracy. Extensive experiments validate that SpaceDrive achieves state-of-the-art open-loop performance on the nuScenes dataset and the second-best Driving Score of 78.02 on the Bench2Drive closed-loop benchmark over existing VLM-based methods. Code is available at: https://github.com/zhenghao2519/SpaceDrive.

cs.CV↗

arXiv2025

"Mapping What I Feel": Understanding Affective Geovisualization Design Through the Lens of People-Place Relationships

Affective visualization design is an emerging research direction focused on communicating and influencing emotion through visualization. However, as revealed by previous research, this area is highly interdisciplinary and involves theories and practices from diverse fields and disciplines, thus awaiting analysis from more fine-grained angles. To address this need, this work focuses on a pioneering and relatively mature sub-area, affective geovisualization design, to further the research in this direction and provide more domain-specific insights. Through an analysis of a curated corpus of affective geovisualization designs using the Person-Process-Place (PPP) model from geographic theory, we derived a design taxonomy that characterizes a variety of methods for eliciting and enhancing emotions through geographic visualization. We also identified four underlying high-level design paradigms of affective geovisualization design (e.g., computational, anthropomorphic) that guide distinct approaches to linking geographic information with human experience. By extending existing affective visualization design frameworks with geographic specificity, we provide additional design examples, domain-specific analyses, and insights to guide future research and practices in this underexplored yet highly innovative domain.

cs.HC↗

arXiv2025

Unveiling the Visual Rhetoric of Persuasive Cartography: A Case Study of the Design of Octopus Maps

When designed deliberately, data visualizations can become powerful persuasive tools, influencing viewers' opinions, values, and actions. While researchers have begun studying this issue (e.g., to evaluate the effects of persuasive visualization), we argue that a fundamental mechanism of persuasion resides in rhetorical construction, a perspective inadequately addressed in current visualization research. To fill this gap, we present a focused analysis of octopus maps, a visual genre that has maintained persuasive power across centuries and achieved significant social impact. Employing rhetorical schema theory, we collected and analyzed 90 octopus maps spanning from the 19th century to contemporary times. We closely examined how octopus maps implement their persuasive intents and constructed a design space that reveals how visual metaphors are strategically constructed and what common rhetorical strategies are applied to components such as maps, octopus imagery, and text. Through the above analysis, we also uncover a set of interesting findings. For instance, contrary to the common perception that octopus maps are primarily a historical phenomenon, our research shows that they remain a lively design convention in today's digital age. Additionally, while most octopus maps stem from Western discourse that views the octopus as an evil symbol, some designs offer alternative interpretations, highlighting the dynamic nature of rhetoric across different sociocultural settings. Lastly, drawing from the lessons provided by octopus maps, we discuss the associated ethical concerns of persuasive visualization.

cs.HC↗

arXiv2025

Prediction of High-Temperature Half Quantum Anomalous Hall Effect in a Semi-magnetic Topological Insulator of MnBi$_2$Te$_4$/Sb$_2$Te$_3$

The classic Thouless-Kohmoto-Nightingale-Nijs theorem dictates that a single electron band of a lattice can only harbor an integer quantum Hall conductance as a multiple of e^2/2h, while recent studies have pointed to the emergence of half quantum anomalous Hall (HQAH) effect, though the underlying microscopic mechanisms remain controversial. Here we propose an ideal platform of MnBi$_2$Te$_4$/Sb$_2$Te$_3$ that allows not only to realize the HQAH effect at much higher temperatures, but also to critically assess the different contributions of the gapped and gapless Dirac bands. We first show that the top surface bands of the Sb$_2$Te$_3$ film become gapped, while the bottom surface bands remain gapless due to proximity coupling with the MnBi$_2$Te$_4$ overlayer. Next we show that such a semi-magnetic topological insulator harbors the HQAH effect at ~20 K, with Cr doping enhancing it to as high as 67 K, driven by large magnetic anisotropy and strong magnetic coupling constants that raise the Curie temperature. Our detailed Berry curvature analysis further helps to reveal that, whereas the gapped surface bands can contribute to the Hall conductance when the chemical potential is tuned to overlap with the bands, these bands have no net contribution when the chemical potential is in the gapped region, leaving the gapless bands to be the sole contributor to the HQAH conductance. Counterintuitively, the part of the gapless bands within the gapped region of the top surface bands have no net contribution, thereby ensuring the plateau nature of the Hall conductance.

cond-mat.mes-hall↗

arXiv2025

TQD-Track: Temporal Query Denoising for 3D Multi-Object Tracking

Query denoising has become a standard training strategy for DETR-based detectors. Denoising queries, initialized by perturbing ground truths, share similarities with track queries in a typical DETR-based Multi-Object Tracking (MOT) method, warranting exploration of their potential synergy. However, query denoising in existing MOT methods is performed only within a single frame, preventing trackers from learning inter-frame temporal association from the denoising process. To address this issue, we propose TQD-Track, a Temporal Query Denoising (TQD) method tailored for MOT. In our method, denoising queries are initialized from ground truths in the previous frame and then propagated into the current frame in the same way as track queries, serving as additional independent data association candidates. These denoising queries carry temporal information and instance-specific feature representations, effectively emulating and augmenting track queries. Moreover, to simulate various real-world MOT challenges for robust tracking, we introduce several corresponding noise types to generate diverse denoising queries. We analyze the impact of our temporal query denoising for two tracking paradigms, tracking-by-attention and alternating detection and association, demonstrating its generalization. Extensive experiments on the nuScenes and Argoverse~2 datasets demonstrate that our approach consistently enhances different MOT baselines, requiring only modifications in the training process. Code and models are available at https://github.com/yutongy98/TQD-Track.

cs.CV↗

arXiv2025

Healing of the edge magnetic island in the island divertor configuration on J-TEXT

The phenomena of island healing and configuration transition induced by high-power electron cyclotron resonance heating (ECRH) have been investigated in the island divertor configuration on the J-TEXT tokamak. Experimental results reveal that the size of the edge open magnetic island with mode number m/n = 3/1 decreases substantially under specific ECRH conditions. This process, referred to as island healing, occurs when ECRH with a power of 500~600 kW is deposited in the plasma core or when 250 kW of ECRH is deposited at r = 0.5 a, where a is the minor radius. The reduction of the island width makes the island divertor ineffective and transition into the limiter configuration. A model incorporating the influence of ECRH on the scrape-off layer (SOL) thermoelectric current is proposed to explain the observed changes in the edge magnetic topology of the island divertor configuration. These findings suggest that ECRH should be deposited at the plasma core with carefully controlled power to ensure the stable and compatible operation of ECRH and the island divertor configuration in tokamaks. The results can provide insights into achieving robust operation of an island divertor in tokamaks.

physics.plasm-ph↗