{"total":37,"items":[{"citing_arxiv_id":"2604.24958","ref_index":5,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Vega-Video: Integrating Video into the Grammar of Graphics","primary_cat":"cs.HC","submitted_at":"2026-04-27T19:56:04+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Vega-Video integrates video into Vega via synchronization, annotation, and transformation classes, using split signals and VOD repurposing for responsive mixed-modality visualizations.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.23830","ref_index":35,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Who Gets to Interpret the Workout? User Tensions with AI-Generated Fitness Feedback","primary_cat":"cs.HC","submitted_at":"2026-04-26T18:24:04+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Qualitative analysis of Reddit discussions reveals four tensions users face with AI-generated fitness feedback, showing resistance to AI that limits personal interpretations of lived experiences.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"Athlete Intelligence, a generative AI feature that automatically pro- duces narrative summaries of users' activities [54]. This feature is marketed as offering athletes personalized insights by aggregating recent activity data into a textual account attached to each workout. In personal informatics (PI) research, this process is termed 'integra- tion': the combination and transformation of data for analysis [35]. Integration is distinct from reflection, which users carry out for themselves as they interpret what data means for their goals and circumstances [42]. By generating narrative summaries automati- cally, Athlete Intelligence performs integration on the user's behalf and presents the results as an interpretation of the workout. Intro- ducing generative AI at this stage, where athletes have traditionally"},{"citing_arxiv_id":"2604.23593","ref_index":104,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"When AI reviews science: Can we trust the referee?","primary_cat":"cs.AI","submitted_at":"2026-04-26T08:03:32+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"AI peer review systems are vulnerable to prompt injections, prestige biases, assertion strength effects, and contextual poisoning, as demonstrated by a new attack taxonomy and causal experiments on real conference submissions.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"50:1097-1179. DOI:10.1162/coli_a_00524 [102] Liu Y ., Deng G., Li Y ., et al. (2023). Prompt injection attack against llm-integrated applications. arXiv preprint. DOI:10.48550/arXiv.2306.05499 [103] Zhou X., Qiang Y ., Zade S.Z., et al. (2023). Hijacking large language models via adversarial in-context learning. arXiv preprint. DOI:10.48550/arXiv.2311.09948 [104] Gong Y ., Chen Z., Chen M., et al. (2025). Topic-fliprag: Topic-orientated adversarial opinion manipulation attacks to retrieval-augmented generation models. arXiv preprint. DOI:10.5555/3766078.3766274 [105] Schwinn L., Dobre D., Günnemann S., et al. (2023). Ad- versarial attacks and defenses in large language models: Old and new threats. arXiv preprint."},{"citing_arxiv_id":"2604.23299","ref_index":38,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Proteus: Shapeshifting Desktop Visualizations for Mobile via Multi-level Intelligent Adaptation","primary_cat":"cs.HC","submitted_at":"2026-04-25T13:28:25+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Proteus uses a multi-level design space and LLM multi-agents to automatically convert desktop visualizations into equivalent mobile versions that preserve readability.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"sage on a smaller canvas. This limitation often results in designs that are technically responsive but cognitively cluttered. 2.2 Automated Visualization Systems Automated visualization design aims to recommend or generate ef- fective charts, given data and tasks, reducing the effort of manual authoring-an effort that remains challenging even for experts [38]. Over the years, prior work has evolved along three major direc- tions. First, rule- and constraint-based systems make design knowl- edge explicit and computable, from foundations such as APT [ 30] to later recommender frameworks (e.g., ShowMe [31], Voyager [51], and Draco [34]), as well as query-driven exploratory approaches [7, 45]. Second, data-driven methods learn charting decisions from"},{"citing_arxiv_id":"2604.22480","ref_index":38,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"AI-based experts' knowledge visualization of cultural heritage: A case study of Terracotta Warriors","primary_cat":"cs.HC","submitted_at":"2026-04-24T12:02:12+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Researchers built a Terracotta Warriors dataset and used AI to analyze and visualize the group attributes collectively.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"responses following disasters. 3.4. Visualization of Multivariate Data Overthepasttwentyyears,avarietyofnewmethodsfor visualizingmultivariatedatahaveemerged,andsomeefforts have been made to survey these approaches [34, 35, 36]. Parallelcoordinatesisanefficientvisualizationmethod[37], oftenusedinthevisualizationofhigh-dimensionalgeometry and multivariate data[38]. Tyagiet al.[39] introduced PC- Expo,aninteractivemethodbasedonmetricsforreordering axes in parallel coordinate displays, which provides an ef- fectivemeansofoptimizingthedisplayofmultivariatedata. <Siyi Li et al.>:Preprint submitted to ElsevierPage 2 of 13 <TWVis> Table 1 The detailed values of each attribute in TW-1087. Attributes Values Definition"},{"citing_arxiv_id":"2604.21830","ref_index":21,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"GFlowState: Visualizing the Training of Generative Flow Networks Beyond the Reward","primary_cat":"cs.LG","submitted_at":"2026-04-23T16:22:55+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"GFlowState introduces interactive visualizations such as trajectory node-link diagrams and transition heatmaps to make GFlowNet training dynamics observable for debugging and quality assessment.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"For example, InstanceFlow [22] and ConfusionFlow [8] monitor the evolution of model class confusion at a local and global level, respectively. While being agnostic to the model type, they are tailored towards classification tasks. For generative tasks, GAN Lab[13], is an educational tool to learn about and experiment with generative adversarial models. For diffusion models,EvolvED[21] introduces evolutionary embeddings to explain the generation process. Reinforcement learning is an ML paradigm more closely related to GFlowNets. Although the learning tasks differ-GFlowNets are gen- erative frameworks, while reinforcement learning involves navigating an agent in an environment-the overall conceptualization of the two approaches is similar: both paradigms involve an environment in which"},{"citing_arxiv_id":"2604.21387","ref_index":27,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"EdgeFormer: local patch-based edge detection transformer on point clouds","primary_cat":"cs.CV","submitted_at":"2026-04-23T07:57:04+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"EdgeFormer converts point cloud edge detection into local-patch point classification with a transformer and reports competitive results against six baselines.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"[25] Ohtake Y, Belyaev A, Seidel HP (2004) Ridge-valley lines on meshes via implicit surface fitting. ACM Trans Graph 23(3):609-612. https://doi.org/10. 1145/1015706.1015768 [26] Lin Y, Wang C, Cheng J, et al (2015) Line segment extraction for large scale unorganized point clouds. ISPRS J Photogramm Remote Sens 102:172-183. https: 19 //doi.org/10.1016/j.isprsjprs.2014.12.027 [27] Stylianou G, Farin G (2004) Crest lines for surface segmentation and flattening. IEEE Trans Vis Comput Graph 10(5):536-544. https://doi.org/10.1109/TVCG. 2004.24 [28] Hildebrandt K, Polthier K, Wardetzky M (2005) Smooth feature lines on sur- face meshes. In: Proceedings of the Third Eurographics Symposium on Geometry Processing. Eurographics Association, Goslar, DEU, SGP '05, pp 85-90"},{"citing_arxiv_id":"2604.20781","ref_index":23,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Designing a Visualization Atlas: Lessons & Reflections from The UK Co-Benefits Atlas for Climate Mitigation","primary_cat":"cs.HC","submitted_at":"2026-04-22T17:09:25+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"The UK Co-Benefits Atlas design process yields a conceptual framework of five driving forces—data, people, stories, context, and the atlas itself—that shape visualization atlas creation at different stages.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"as part of this in order to scaffold the design of future atlases. 2.2 Approaches to Visualization Design A rich body of work exists on visualization methodology, focused in particular, on scaffolding visualization design processes. Some of the most prominent examples include the four key visualization design activities of understanding, ideating, making, and deploying [23]; the classic nine-step process of visualization design studies [37]; and the extension towards data-first design studies [27]. Across these models, task abstraction plays a central role, supported by frameworks such as the nested model [25] and the multi-level typology [4]. While these models are widely applicable, most of these frameworks are built on"},{"citing_arxiv_id":"2604.20759","ref_index":60,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Autark: A Serverless Toolkit for Prototyping Urban Visual Analytics Systems","primary_cat":"cs.HC","submitted_at":"2026-04-22T16:48:24+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Autark is a serverless toolkit that enables rapid prototyping of urban visual analytics systems via domain-aware abstractions and supports more reliable LLM-assisted coding.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"ates complete ParaView [7] scripts for 3D scientific visualization work- flows, and NLI4V olVis [8] enables real-time exploration and semantic editing of volumetric scenes using natural language. More recently, re- searchers have begun leveraging LLMs not only to generate individual visual artifacts but also to orchestrate entire V A systems. LightV A [60] decomposes high-level analytical goals into executable subtasks to facilitate coordinated visualizations, while Data-to-Dashboard [59] au- tomates end-to-end pipelines from raw data to interactive dashboards using modular agents that iterate and reflect on their decisions. A fully general framework, however, capable of assembling V A systems across"},{"citing_arxiv_id":"2604.18801","ref_index":34,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Preserving Clusters in Error-Bounded Lossy Compression of Particle Data","primary_cat":"cs.LG","submitted_at":"2026-04-20T20:10:54+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"A clustering-aware correction algorithm using spatial partitioning and projected gradient descent preserves single-linkage clusters in lossy-compressed particle data while keeping competitive compression ratios.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"decompressed output by applying \"edits\" to restore specific features while strictly adhering to the original pointwise error bounds. For example, to preserve power spectra, FFCz [8] derives edits by alternating projections between spatial and fre- quency constraints. To preserve scalar field topology, MSz [33] edits decompressed data to maintain Morse-Smale segmenta- tions, while Gorski et al. [34] focus on contour tree accu- racy. However, unlike these existing methods which focus on regular-grid scalar fields, preserving the integrity of single- linkage clusters in particle data remains an open challenge, as it requires managing the high sensitivity of distance to small coordinate changes near the linking threshold. B. Single-linkage clustering"},{"citing_arxiv_id":"2604.15813","ref_index":60,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Exploring Agentic Visual Analytics: A Co-Evolutionary Framework of Roles and Workflows","primary_cat":"cs.DB","submitted_at":"2026-04-17T08:11:39+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"A survey of 55 agentic VA systems proposes a co-evolutionary framework defining four agent roles (PLANNER, CREATOR, REVIEWER, CONTEXT MANAGER) mapped to visual analytics pipeline stages along with design guidelines.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"DynaVis [55] 2024 2 Data Formulator [59] 2024 2 DASH [3] 2024 2 Bavisitter [8] 2024 3 NL4DV-LLM [41] 2024 3 FathomGPT [22] 2024 3 MatPlotAgent [66] 2024 3 A V A [29] 2024 3 DataVisT5 [57] 2025 1 HARVis [46] 2025 1 SpeechVisNet [68] 2025 1 CoML4Vis [5] 2025 1 NL4DV-Stylist [20] 2025 2 Prompt4Vis [26] 2025 2 SUPQA [17] 2025 2 Aggarwal et al. [1] 2025 2 Jupybara [60] 2025 2 Data Formulator 2 [58] 2025 2 Pluto [50] 2025 2 RECITKIT [42] 2025 2 Step-NL2VIS [30] 2025 2 InterChat [4] 2025 2 Dataweaver [12] 2025 2 ChartGPT [54] 2025 2 VizTA [62] 2025 2 Shao et al. [43] 2025 2 nvAgent [37] 2025 3 C2 [24] 2025 3 Plume [27] 2025 3 DeepVis [47] 2025 3 Text2Vis [40] 2025 3 DataNarrative [19] 2025 3 CoDA [7] 2025 3 Gyarmati et al."},{"citing_arxiv_id":"2604.26967","ref_index":18,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Literate Execution","primary_cat":"cs.PL","submitted_at":"2026-04-17T06:55:44+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Literate execution treats documentation and visualizations as dynamic, computable parts of program execution via provenance tracking, inverting traditional literate programming to make programs more explorable.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"and Jupyter development team. 2016. Jupyter Notebooks - a Publishing Format for Reproducible Computational Workflows. InPositioning and Power in Academic Publishing: Players, Agents and Agendas, Fernando Loizides and Birgit Scmidt (Eds.). IOS Press, 87-90. [17] Donald E. Knuth. 1984. Literate Programming.Comput. J.27, 2 (May 1984), 97-111. doi:10.1093/comjnl/27.2.97 [18] Brittany Kondo and Christopher Collins. 2014. DimpVis: Exploring Time-Varying Information Visualizations by Direct Manipulation.IEEE Transactions on Visualization and Computer Graphics20, 12 (2014), 2003-2012. doi:10.1109/TVCG. 2014.2346250 [19] Sorin Lerner. 2020. Projection Boxes: On-the-fly Reconfigurable Visualization for Live Programming. InProceedings of"},{"citing_arxiv_id":"2604.14365","ref_index":41,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Interactive Exploration of Large-scale Streamlines of Vector Fields via a Curve Segment Neighborhood Graph","primary_cat":"cs.CG","submitted_at":"2026-04-15T19:27:31+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A web system uses a Curve Segment Neighborhood Graph to support interactive community detection, force-directed layouts, and adjacency matrix views for exploring hundreds of thousands of streamlines in real time.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"While they focused on continuous mani- fold representation of whole streamlines, our work uses a dis- crete graph-based representation (CSNG) modeling topological neighborhoods at the segment level, enabling local flow com- munity detection through modularity optimization. Pattern search identifies curves or segments similar to user- specified references. Wang et al. [41, 42, 7] developed pattern identification methodologies for vector field data. Lu et al. [5] introduced distribution-based streamline characterization. Tao et al. [43, 6] encoded streamline characteristics into charac- ter strings to facilitate pattern search. Pattern search strategy requires user-specified references as input and may miss impor- tant patterns that the user has no knowledge of."},{"citing_arxiv_id":"2604.11172","ref_index":23,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"NeuVolEx: Implicit Neural Features for Volume Exploration","primary_cat":"cs.GR","submitted_at":"2026-04-13T08:30:39+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"NeuVolEx extracts robust spatial features from INR training via a structural encoder and multi-task scheme to enable accurate ROI classification with limited supervision and unsupervised viewpoint clustering in volume exploration.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"Viewpoint Sampling Multi-class Probabilistic Vo lu m e R e n d e re r User-guided V oxel Classification Set-cover Problem Solver Fig. 2: Overview of our NeuVolEx approach. mapping allowed users to manipulate appearance of visible ROIs within DVR images. Yao et al. [32] introduced ViSNeRF, which extends NeRF to time-varying volume series. Tang et al. [23] augmented this coordinate-optical mapping with style transfer. As such, the potential of INRs for fundamental volume exploration tasks, such as image-based TF design or viewpoint recommendation, remains largely overlooked. Moreover, prior work has primarily focused on INR outputs, while the feature representations learned during INR training have received"},{"citing_arxiv_id":"2604.10237","ref_index":15,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Glide-in-Place: Foot-Steered Differential-Drive for Hands-Free VR Locomotion","primary_cat":"cs.HC","submitted_at":"2026-04-11T14:39:34+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Glide-in-Place uses per-foot pressure sensing for continuous differential-drive VR locomotion, outperforming seated walking-in-place on speed and fatigue while matching joystick performance on simulator sickness in a 16-person study.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.10008","ref_index":29,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Raiven: LLM-Based Visualization Authoring via Domain-Specific Language Mediation","primary_cat":"cs.HC","submitted_at":"2026-04-11T03:33:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Raiven mediates LLM visualization authoring via a formally defined DSL that unifies scientific and information visualization, producing deterministic, verifiable code from metadata-only inputs.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"System Year NL DSLS I FlowSense [44] 2020 Chat2Vis [23] 2023 LIDA [8] 2023 ChatVis [25] 2024 Omega [32] 2024 ParaView-MCP [21] 2025 Diderot [6] 2012 Diderot Vivaldi [7] 2014 Vivaldi ViSlang [30] 2014 Vega-Lite [33] 2017 Vega-Lite Shih et al. [36] 2019 DXR [37] 2019 Scholz [34] 2021 Harth et al. [10] 2023 GoFish [29] 2026 GoFish Gosling [17] 2026 Gosling NL4DV [28] 2021 Vega-Lite NMT2Vis [22] 2022 VegaZero NL2Viz [43] 2022 FlowNL [13] 2023 ChatModelling [15] 2024 Y AC [19] 2025 ChartGPT [38] 2025 NLI4V olVis [1] 2026 VegaChat [12] 2026 Vega-Lite Raiven 2026 RaivenDSL library provides programmatic access to rendering primitives but im-"},{"citing_arxiv_id":"2604.09241","ref_index":25,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"LandSAR: Visceralizing Landslide Data for Enhanced Situational Awareness in Immersive Analytics","primary_cat":"cs.HC","submitted_at":"2026-04-10T11:55:13+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"LandSAR integrates real-time landslide simulations, visualizations, and 3D-printed tangible terrain models to improve situational awareness and engagement for analysts.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"function as real-world sliders. Tonget al.[63] extensively studied the interaction design space of printed data visualizations. These interfaces, however, are created without real-world meaning and are entirely de- tached from the application, similar to a mouse and keyboard [60, 64]. While these tools enable physical interactions, they still lack physical contexts. Flecket al.[25] noted that the main differences between immersive and situated analytics lie in the absence of physical context, interaction, and embedded perception of virtual content. To incorporate spatial meaning into interfaces, Satriadiet al.[56] examined scale models and explored view arrangement and chart layout for visualizing multivariate data and multiple charts."},{"citing_arxiv_id":"2604.09134","ref_index":25,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Enhance Comprehension of Over-the-Counter Drug Instructions for the General Public and Medical Professionals through Visualization Design","primary_cat":"cs.HC","submitted_at":"2026-04-10T09:14:51+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Tailored visualizations for OTC drug instructions improve response time and usability over text-only formats for both lay users and professionals, supported by a new taxonomy and generalizable design workflow.","context_count":1,"top_context_role":"background","top_context_polarity":"support","context_text":"Our study combines three research areas: pharmaceutical in- formation comprehension, visualization for medical communi- cation, and visualization design studies. The following gives an overview of recent works in these areas. 3.1. Pharmaceutical Information Comprehension Researchers explore the specific support people need to make sense of visualizations, aiming to improve visualization de- sign [25], which may enhance comprehension. Pictographic visualizations could improve user engagement and comprehen- sion [26, 27]. A systematic review [28] concludes that it is beneficial to include pictures in medical communication, and icons/pictograms with less text may be most helpful. Many early studies regarding pharmaceutical information comprehension focus on pictograms/icons like the United"},{"citing_arxiv_id":"2604.07989","ref_index":21,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Show Me the Infographic I Imagine: Intent-Aware Infographic Retrieval for Authoring Support","primary_cat":"cs.IR","submitted_at":"2026-04-09T08:58:59+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Presents a new retrieval system that enriches user queries with an intent taxonomy to improve matching of natural language descriptions to infographic designs and support authoring.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"Graphics, 13(6):1137-1144, 2007. doi: 10.1109/TVCG.2007.70594 2 [20] D. Moritz, C. Wang, G. L. Nelson, H. Lin, A. M. Smith, B. Howe, and J. Heer. Formalizing visualization design knowledge as constraints: Ac- tionable and extensible models in draco.IEEE Transactions on Visualiza- tion and Computer Graphics, 25(1):438-448, 2019. doi: 10.1109/TVCG. 2018.2865240 2 [21] A. Narechania, A. Srinivasan, and J. Stasko. NL4DV: A toolkit for gener- ating analytic specifications for data visualization from natural language queries.IEEE Transactions on Visualization and Computer Graphics, 27(2):369-379, 2021. doi: 10.1109/TVCG.2020.3030378 2 [22] H. N. Nguyen and N. Gehlenborg. Safire: Similarity framework for visualization retrieval."},{"citing_arxiv_id":"2604.06358","ref_index":10,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"GS-Surrogate: Deformable Gaussian Splatting for Parameter Space Exploration of Ensemble Simulations","primary_cat":"cs.GR","submitted_at":"2026-04-07T18:37:15+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"GS-Surrogate creates a canonical Gaussian field that is sequentially deformed by simulation parameters to enable real-time, controllable 3D exploration of ensemble data while separating simulation variations from visualization adjustments.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"reintroducing significant I/O and storage overhead. Existing approaches for parameter-space exploration of ensemble simulations fall into two main categories. The first one relies on stan- dard high-dimensional data visualization techniques applied directly to collected ensemble inputs and outputs. Parallel coordinates [27, 33], scatter plots [25, 28], radial plots [10, 14], glyphs [9], and matrix-based views [29] have all been used to analyze relationships across ensemble members. A fundamental limitation shared by all these methods is that analysis remains confined to parameter configurations that were explicitly simulated. The second category, including our GS-Surrogate, uses surrogate models to predict outcomes at new, unsampled parameter"},{"citing_arxiv_id":"2604.05200","ref_index":1,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Investigating Ethical Data Communication with Purrsuasion: An Educational Game about Negotiated Data Disclosure","primary_cat":"cs.HC","submitted_at":"2026-04-06T21:57:07+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Purrsuasion is a negotiation game that surfaces satisficing and intent-attribution difficulties when students practice ethical data disclosure under real constraints.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"fer HCI researchers a way to study how people reason under structured roles, partial information, and competing goals. Visualization research therefore stands to benefit from studying communication games: in- teractions in which participants make strategic choices about how and what to communicate through visualizations. Game-based approaches have already been used productively in visualization contexts [1, 2]. We draw on this tradition by presenting a game inspired in part by negotiation activities common in business school programs. Observing disclosure through a game is methodologically useful because it turns otherwise tacit design trade-offs into visible, consequential actions that can be compared across players, rounds, and roles. InPurrsuasion,"},{"citing_arxiv_id":"2604.03520","ref_index":53,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"SwEYEpinch: Exploring Intuitive, Efficient Text Entry for Extended Reality via Eye and Hand Tracking","primary_cat":"cs.HC","submitted_at":"2026-04-03T23:53:43+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"SwEYEpinch uses gaze swiping plus a held pinch gesture to reach 64.7 WPM in XR text entry after practice, outperforming sequential key selection and prior gaze-swipe methods in user studies.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"of the SIGCHI conference on Human factors in computing systems. 113-120. [51] Robyn Speer. 2022. rspeer/wordfreq: v3.0. (Sept. 2022). DOI:http://dx.doi.org/10. 5281/zenodo.7199437 [52] Marco Speicher, Anna Maria Feit, Pascal Ziegler, and Antonio Krüger. 2018. Selection-based text entry in virtual reality. InProceedings of the 2018 CHI Con- ference on Human Factors in Computing Systems. 1-13. [53] Paul Streli, Jiaxi Jiang, Andreas Rene Fender, Manuel Meier, Hugo Romat, and Christian Holz. 2022. TapType: Ten-finger text entry on everyday surfaces via Bayesian inference. InProceedings of the 2022 CHI Conference on Human Factors in Computing Systems. 1-16. [54] Uta Wagner, Andreas Asferg Jacobsen, Tiare Feuchtner, Hans Gellersen, and Ken Pfeuffer."},{"citing_arxiv_id":"2604.03406","ref_index":50,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"SASAV: Self-Directed Agent for Scientific Analysis and Visualization","primary_cat":"cs.GR","submitted_at":"2026-04-03T19:09:38+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"SASAV introduces the first fully autonomous multi-agent system for scientific data analysis and visualization that operates without external prompting or human-in-the-loop feedback.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.02535","ref_index":61,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"A Spectral Framework for Multi-Scale Nonlinear Dimensionality Reduction","primary_cat":"cs.LG","submitted_at":"2026-04-02T21:39:52+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A spectral framework for nonlinear DR uses spectral bases plus cross-entropy optimization to create multi-scale embeddings that preserve both global manifold geometry and local neighborhoods while supporting graph-frequency analysis.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"is then optimized to minimize the cross-entropy between the high- dimensional similarities wi,j and the low-dimensional similarities qi,j . In practice, UMAP symmetrizes the fuzzy graph and applies stochastic optimization with negative sampling to reduce calculation time. Cross-entropy optimization is also used in other DR methods such as LargeVis [61], due to its flexibility and computational efficiency. Our framework adopts this optimization strategy but applies it to a spectral subspace. Specifically, we optimize an M′-dimensional embedding of a higher-dimensional spectral subspace (e.g., S=20 ), enabling a balance between expressiveness and explainability of the resulting embedding. 2.4 Analysis and Understanding of DR Results"},{"citing_arxiv_id":"2604.16355","ref_index":36,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"A Multi-Technique Approach for Improving Summary Polar Diagrams","primary_cat":"cs.HC","submitted_at":"2026-03-17T17:34:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":3.0,"formal_verification":"none","one_line_summary":"Combining overview+detail, aggregation, filtering, linking, and small multiples makes summary polar diagrams clearer and more usable for discovering data relationships.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.15342","ref_index":50,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"SuperProvenanceWidgets: Tracking and Visualizing Analytic Provenance Across UI Control Elements","primary_cat":"cs.HC","submitted_at":"2026-03-13T16:06:26+00:00","verdict":"ACCEPT","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"SuperProvenanceWidgets adds cross-control provenance tracking to visualize how users interact with multiple UI elements over time, demonstrated through usage scenarios and a developer usability assessment.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.09598","ref_index":2,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Visualization Retrieval for Data Literacy: Position Paper","primary_cat":"cs.HC","submitted_at":"2026-03-06T16:08:30+00:00","verdict":"CONDITIONAL","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Visualization retrieval systems can transform static collections of visualizations into dynamic, inquiry-based environments that support design exploration, data consumption, and resource management for data literacy education.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.15320","ref_index":6,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Towards Measuring Interactive Visualization Abilities: Connecting With Existing Literacies and Assessments","primary_cat":"cs.HC","submitted_at":"2026-03-02T22:38:43+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"The paper proposes and compares approaches for assessing interactive visualization abilities by linking them to existing literacy concepts and assessments.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.06187","ref_index":20,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Skin-Deep Bias: How Avatar Appearances Shape Perceptions of AI Hiring","primary_cat":"cs.HC","submitted_at":"2026-02-16T14:13:48+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Racial mismatch between applicant and AI avatar increased perceived ethnic bias, while sharing only one identity trait lowered fairness ratings compared to full or no match.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2509.19182","ref_index":10,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"YAC: Bridging Natural Language and Interactive Visual Exploration with Generative AI for Biomedical Data Discovery","primary_cat":"cs.HC","submitted_at":"2025-09-23T15:57:42+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"YAC is a prototype system that uses a tool-calling multi-agent architecture to translate natural language into linked interactive visualizations and filters for biomedical data, with user-adjustable structured output and a domain-expert user study.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2507.22051","ref_index":68,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"DataSway: Vivifying Metaphoric Visualization with Animation Clip Generation and Coordination","primary_cat":"cs.HC","submitted_at":"2025-07-29T17:55:41+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"DataSway supports creation of semantically aligned animations for metaphoric data visualizations by generating clips via VLMs and coordinating timelines based on entity order, attributes, layout, or randomness.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2506.14611","ref_index":68,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Exploring MLLMs Perception of Network Visualization Principles","primary_cat":"cs.HC","submitted_at":"2025-06-17T15:10:01+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"MLLMs given the same instructions as human participants achieve expert-level performance on perceiving stress in network visualizations and rely on similar visual proxies.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2504.02204","ref_index":51,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Characterizing Creativity in Data Visualization: Reflections and Future Directions","primary_cat":"cs.HC","submitted_at":"2025-04-03T01:31:09+00:00","verdict":"ACCEPT","verdict_confidence":"MODERATE","novelty_score":4.0,"formal_verification":"none","one_line_summary":"A systematic review and interview study characterize creativity in visualization design, finding that design processes are undervalued compared to final artifacts with ideation as a universal bottleneck.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2408.02679","ref_index":43,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Visual Analysis of Multi-outcome Causal Graphs","primary_cat":"cs.LG","submitted_at":"2024-07-31T08:44:34+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Introduces progressive visualization for comparing causal discovery algorithms and comparative graph layouts for analyzing multi-outcome causal graphs in healthcare.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2405.15491","ref_index":22,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"GSDeformer: Direct, Real-time and Extensible Cage-based Deformation for 3D Gaussian Splatting","primary_cat":"cs.CV","submitted_at":"2024-05-24T12:16:28+00:00","verdict":null,"verdict_confidence":null,"novelty_score":null,"formal_verification":null,"one_line_summary":null,"context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2306.06043","ref_index":92,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"A Qualitative Analysis of Common Practices in Annotations: A Taxonomy and Design Space","primary_cat":"cs.HC","submitted_at":"2023-06-09T17:12:01+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A taxonomy and design space for chart annotations synthesized from qualitative coding of 1,800 static real-world examples.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1907.08325","ref_index":34,"ref_count":1,"confidence":0.5,"is_internal_anchor":false,"paper_title":"Scalable Topological Data Analysis and Visualization for Evaluating Data-Driven Models in Scientific Applications","primary_cat":"cs.LG","submitted_at":"2019-07-19T00:37:39+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A scalable framework combining streaming graphs, topology computation, and topology-aware datacubes enables interactive analysis of high-dimensional functions in scientific ML applications.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null}],"limit":50,"offset":0}