Papers for

data visualization developers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

LLM judges struggle to match human visualization preferences

Evaluating Human-LLM Alignment on Visualization Preference

Abstract: There are often multiple valid visualizations, spanning different chart types, visual encodings, and aesthetic choices, that can all support a user's goal of analyzing a dataset. Determining which of these candidate visualizations is most appropriate, however, remains challenging. Prior work has leveraged expert-curated taxonomies to guide visualization design and subsequent quality evaluation, a practical evaluation model must also account for human-centric factors. Latest visualization tools integrate Large Language Models (LLMs) to generate and refine visualizations, using LLM-as-a-judge to provide design feedback. Without a deeper understanding of whether, and in what ways, human and LLM perceptions of visualizations align, there is a real risk of deploying low-quality agents that do not reflect human judgment. We study this question through a pilot experiment - we formulate the problem as a preference task over curated chart pairs and compare the judgments of state-of-the-art vision models with those of a panel of human annotators. We establish a cautionary tale: such judges risk filtering out informative, compelling displays. We find that frontier models collapse human subjectivity, a rich signal that reveals contrasting design choices, into near-deterministic verdicts.

Wed 4 NovHuman-Computer Interaction
The gist
Choosing the best way to show data with charts is tricky since many good options exist. The researchers looked at how well AI models agree with people when picking preferred visualizations. They found that AI tends to give very firm answers, ignoring the natural differences in human opinions. This means current AI may miss some creative or useful designs people like.
Open → 2611.06272v1

New color maps improve spherical maps by matching color to distance

SpheriColor: Colormaps for Spherical Geospatial Input Topographies

Abstract: Multivariate geospatial data visualization often relies on multiple coordinated views, where color can be used to either link views or encode data attributes. Encoding spatial locations through color can reveal patterns in non-spatial visualizations, yet most applications of colormaps focus on high-dimensional attribute encodings instead. While 2D colormaps have been studied extensively, color encodings designed for spherical geospatial data have received much less attention, even though all locations lie on a sphere. To address this, we propose SpheriColor, a colormap generation approach for mapping geospatial references guided by the design goals of distance preservation and colorspace exploitation. We project a geospatial distance function into a perceptually linear colorspace, followed by two different gamut-constrained optimization strategies. We perform a quantitative evaluation using both real-world and synthetic datasets, demonstrating superior performance over 2D colormaps and HSLuv double cone encodings. The simplex-based optimization excels at distance preservation, whereas the ray-based approach provides better colorspace exploitation.

Thu 8 OctHuman-Computer Interaction
The gist
When showing data on a globe, colors can help people understand how far apart places are. Most color methods work well on flat maps but not on spherical ones like the Earth. The authors created a way to pick colors for maps on spheres that keep distances clear and use colors better. Their tests show their methods work better than older ways, helping us see spatial data more clearly.
Open → 2610.11470v1

ViaCross crossing gestures improve complex object and attribute selection

viaCross: A Pen-Based Technique for Selecting Objects and Attributes of Interest using Crossing-Based Gestures

Abstract: Crossing-based selection is well-studied, yet past research has overlooked multi-object and attribute selection with crossing gestures. We present viaCross, a pen-based technique in which crossing strokes define attribute constraints, paired with attribute widgets that visualize and allow subsequent edits to these constraints. A controlled study compared viaCross with traditional WIMP filter panels. Results show that viaCross was more efficient for complex object-level selections, requiring fewer strokes and maintaining efficiency as task difficulty increased. WIMP panels were faster and more intuitive for simple attribute-based selections, but participants encountered repeated deselections and occasional misclicks. viaCross matched panel performance for creating attribute selections but was slower during refinement, as widgets had to be recreated, indicating the need for persistent widgets or hybrid support with a WIMP-style interface. Overall, viaCross demonstrates that crossing-based gestures can provide an efficient and scalable approach to multi-object and attribute selection.

Fri 2 OctHuman-Computer Interaction
The gist
Selecting multiple objects and their specific properties can be tricky with traditional mouse and menu tools. The authors introduce viaCross, a pen-based method where you draw lines that cross objects and set conditions on their features. Their study found that viaCross worked better than normal menus for tough tasks requiring many selections, though menus were faster for simple choices. The study also revealed some limits in refining selections with viaCross, suggesting future improvements combining both approaches. Overall, it shows that drawing crossing gestures with a pen can be a powerful way to select many items and their details.
Open → 2610.03995v1

Large language models struggle with current chart-making languages

Where LLMs Fail with Visualization DSLs

Abstract: As LLMs take up the role of authoring charts using visualization domain-specific languages (DSLs), the human constraints that shaped those languages may no longer apply, as what is easy for a person is not necessarily easy for a model. To understand how LLMs might work better with DSLs, we explore where and how they fail with current DSL designs. We evaluate 10 JSON-style visualization DSLs with 41 tasks across 3 LLMs, then assess the generated specifications with JSON and rendering checks, and qualitative coding of failed cases. Analyzing how this specification generation process fails, we identify four recurring failure patterns, link each to specific DSL features, and discuss design considerations for future DSL designs.

Thu 1 OctHuman-Computer InteractionComputation and Language
The gist
Making charts with special computer languages is often designed to be easy for humans, but these languages can be tricky for large language models (LLMs). The authors tested three popular LLMs on 10 different chart-making languages with many tasks to see where the models make mistakes. They found four common types of errors and connected these problems to certain features of the languages. This helps understand how to design better chart languages that computers can use more effectively.
Open → 2610.01873v1

InfoAgent tracks and fixes facts in editable infographics

InfoAgent: Traceable Generation and Repair of Evidence-Grounded Infographics

Abstract: Reliable infographic generation requires facts, symbols, and visual relations to remain consistent through rendering and revision. Correcting one element also requires tracking its supporting evidence and the dependencies affected by the change. We present \textbf{InfoAgent}, a training-free framework for \emph{evidence-bound visual-symbolic program synthesis}. Its Infographic Visual Description (IVD) records factual payloads, evidence provenance, execution routes, and verification obligations in a typed dependency graph. Retrieved design priors guide compilation, and layered execution combines raster synthesis with editable symbolic and binding objects while retaining their traces. Dependency-aware repair localizes corrections, rechecks affected dependencies, and requires protected obligations to remain satisfied under the declared checkers. Unresolved obligations remain explicit. On IGenBench, InfoAgent achieves 93.0 Q-ACC and 59.0 I-ACC. We also introduce InfoGraphicBench-Evidence, where complete-checklist pass rates on 200 test requests increase from 21.5\% for Same-IVD Prompt to 23.5\% for the initial layered output and 28.5\% after repair, using the same evidence and initial IVD. On 120 audited repair cases, localized repair edits 12.4\% of the canvas on average, compared with 67.3\% for global regeneration.

Wed 30 SeptComputer Vision and Pattern Recognition
The gist
Creating accurate infographics is hard because changing one fact or image part means you have to update other connected pieces correctly. The authors built InfoAgent, a tool that keeps track of every fact and its proof as a connected map, so it can generate infographics that stay consistent. When something needs fixing, InfoAgent knows which parts to change and checks that the overall infographic remains correct. This makes fixing errors faster and keeps the design and evidence linked throughout.
Open → 2609.39380v1

Models use linear number representations despite curved geometry in tasks

When Models Don't Manipulate Manifolds: The Geometry of a Comparison Task

Abstract: One of the current premises of mechanistic interpretability research is that detailed accounts of the geometry of neural network representations can tell us how models perform computations, and how to effectively intervene on them. While low dimensional manifolds have been observed for multiple concepts in the literature (e.g. numbers encoded on helices, days of the week on a circle, ...), with structure believed to reflect properties of data and tasks, the extent to which models rely on them for computation, and how they manipulate them, remains unclear. We characterize precisely the geometry of computation in a number-comparison task, as an abstraction of comparison for decision making, and how models utilize geometry in an elegant fashion to implement it. Specifically, we study the causal geometry of number comparison in Qwen2.5-7B-Instruct, a capable and widely studied open-weight model, and find Qwen largely uses linear representations of numbers despite the presence of curved geometry. To compare two numbers, the model first encodes each number along a vector and adds the two representations using attention and the residual connection, bringing them into a shared space in the residual stream. Then, the model uses MLP neurons to compare the pair of numbers on local regions in this shared space, which correspond to smaller intervals of input numbers, and combines these to obtain the position of the maximum. In fact, this reliance on linear representations for comparison also persists when the model compares three numbers. Our findings demonstrate that the manifold hypothesis can co-exist with linear representations: while concepts that are ordered may have manifold structure in representations, the model may use an underlying linear structure of the concept in certain computations.

Tue 29 SeptMachine LearningArtificial IntelligenceComputation and Language
The gist
This paper looks at how AI models compare numbers inside their 'brains.' The authors studied a popular language model and found that although numbers can be represented in curved shapes, the model actually treats them in simpler, straight-line ways for comparing tasks. It first encodes each number as a line, combines them, and then compares parts of the combined numbers to decide which is bigger. This means AI can mix complex shapes and simple lines depending on what helps the task.
Open → 2609.37680v1

Interactive system improves chart making from imperfect text requests

VisInteract: Towards Dynamic Interactive Text-to-Visualization under Imperfect Queries

Abstract: Real-world visualization requests are routinely ambiguous, incomplete, or factually incorrect, yet existing Text-to-Visualization (Text-to-Vis) systems assume well-specified inputs and produce charts in a single pass. When queries are imperfect, a system must \emph{interact} with the user to recover the true intent, but no benchmark or method supports this dynamic process. We introduce \textbf{VisInteract}, a new paradigm that reframes Text-to-Vis as interaction-driven intent recovery, and \textbf{VisInteract-Bench}, to our knowledge, that is the first benchmark for dynamic interactive Text-to-Vis, featuring controlled imperfection injection, a leakage-controlled User Agent for realistic multi-turn feedback, and dual-perspective (code and chart) automated evaluation. On the algorithmic side, we propose \textbf{Vis-MCTS}, a Monte Carlo Tree Search (MCTS) enhanced method, introducing improvements over classical MCTS, that \emph{Progressive Widening} to tame the unbounded tool-argument space in tree search, \emph{cross-rollout information sharing} so clarifications and critiques benefit the entire search tree, and \emph{Dimension-Aware Reward Decomposition} that routes scalar user feedback along data-fidelity, visual-design, and intent-alignment dimensions to resolve credit assignment across heterogeneous actions. Extensive Experiments across two LLM backbones show that Vis-MCTS consistently outperforms all Text-to-Vis baselines, improving end-to-end task success by $13.40\%$--$16.27\%$ over the strongest interactive baseline and by more than $5\times$ over non-interactive ones.

Mon 14 SeptArtificial Intelligence
The gist
People often ask computers to create charts from text instructions, but these requests can be unclear or incorrect. The authors propose VisInteract, which helps the system talk back and forth with users to better understand what chart they really want. They also created a test setup to see how well this interactive process works when the instructions are imperfect. Their improved method, Vis-MCTS, uses smart search techniques to ask clarifying questions and learn from feedback, leading to much better final charts. This approach outperforms previous non-interactive and interactive methods by a wide margin.
Open → 2609.15182v1

Ggaction models chart creation as a chain of graphical actions

ggaction: A Grammar of Graphical Actions

Abstract: A chart may be declarative; authoring it is not. Visualization grammars often describe charts as finished specifications, whereas people construct them through a sequence of authoring actions. This mismatch can make visualization code difficult for humans to interpret and for machines to generate from human intent. ggaction addresses this gap by modeling the chart authoring process itself. In ggaction, individual authoring actions are abstracted as functions, and the authoring process is expressed as a chain of these functions. This representation more closely aligns chart designers' authoring intent with code specifications, making the code easily understandable to both humans and machines, including language models. Through a series of evaluations, we show that ggaction is sufficiently expressive to capture common chart authoring intents and outperforms widely used visualization grammars, including Vega-Lite and ggplot2, in both human and machine interpretability. ggaction is available at github.com/ggaction/ggaction.

Sun 13 SeptHuman-Computer InteractionGraphics
The gist
Making charts on a computer is usually described as a finished product, but actually building one happens step-by-step. The authors introduce ggaction, which treats each step as a separate action in a chain, making the process easier for people and machines to understand. This approach better matches how people think about making charts compared to other tools. Tests show ggaction is good at capturing common ways people build charts and is easier to interpret than popular systems like Vega-Lite and ggplot2.
Open → 2609.14353v1

When sketch input improves large language model chart annotations

AnnoSketch: Evaluating and Collecting Human Sketches for MLLM-assisted Chart Annotation

Abstract: As multimodal large language models (MLLMs) support a growing range of input modalities, increasing work explores how to incorporate rough sketches to convey user intent. For annotated chart generation, it remains unclear what annotation sketches people provide and when such visual input helps MLLMs generate more useful annotations. In this study, we examine when sketch input is useful for MLLM-generated chart annotations across variation in chart type and caption type. In addition, we qualitatively analyze participants' explanations of their output preferences to characterize what made generated annotations more or less helpful. To further document participants' annotation sketches, we present AnnoSketch, comprising 1,600 annotation sketches collected across 160 chart-caption pairs from the conditions in which sketch guidance proved most beneficial, together with participants' annotation intents, perceived comprehension difficulty, and self-reported expressive limitations. We also label these sketches with structured metadata describing how each sketch relates to its caption and how participants express annotations through visual marks. Together, our study and AnnoSketch help determine when to solicit sketch input and provide empirical source for how people sketch chart annotations to support captions. The dataset and supplemental materials are available in our OSF repository.

Sun 13 SeptHuman-Computer Interaction
The gist
It can be hard for computers to explain charts clearly by themselves. This study looks at how rough sketches provided by people can help large language models create better descriptions and notes for charts. The authors collected a large set of sketch examples that show how people annotate charts visually and analyzed when these sketches actually improve the computer’s output. Their work helps understand when to ask users for sketches and provides data on how people draw them to support chart explanations.
Open → 2609.14289v1

Vision language models identify key regions to read bar chart values accurately

Inside VLM Chart Reading: Tracing Value Reading from Vertical Bar Charts Across Space and Depth

Abstract: Vision--language models (VLMs) can answer chart questions accurately, but output accuracy does not show how they combine the evidence needed to recover an exact value. We study vertical-bar value reading with controlled counterfactual activation patching in Qwen2.5VL-7B-Instruct and InternVL3.5-8B. The study connects three analyses: (1) The single-factor results show that the changed bar-top region restores much more answer preference than the unchanged bar body, despite containing fewer visual tokens. Legend- and series-related states also lose local recoverability earlier than bar-geometry and axis-scale states. (2) In the handoff analysis, restoration shifts from visual legend regions in early layers to prompt-series positions in middle layers. Resetting the prompt-series state selectively reduces legend-source rescue, supporting its role as a partial mediator. (3) In the factorial analysis, both models can use geometry and scale states from separate donors to favor the combined target. InternVL performs similarly when the states come from separate donors or one image, while Qwen shows lower restoration for separate donors which suggests higher context sensitivity. Together, these results provide preliminary causal evidence for localizing the internal computations that support exact bar-value reading.

Sat 12 SeptComputation and Language
The gist
Some advanced AI models can answer questions about bar charts, but it wasn’t clear exactly how they figure out the exact numbers shown. The authors studied two such models and found that the top part of each bar in a chart plays a bigger role than the rest of the bar for reading values. They also discovered that these models shift where they get information from as they process the chart, moving from the legend in early stages to other parts of the input later on. This helps show how the models combine different pieces of the chart to determine precise values.
Open → 2609.13745v1

Interactive authoring tools improve science data visualization and analysis

"Here Be Sharks!": Enhancing Scientific Communication and Analysis through Authoring Interactivity

Abstract: We report on a case study for designing authoring environments for interactive visualizations to enhance scientific work. We conducted a workshop and prototype review with a group of marine biologists. When it came to visualizing their data, participants identified challenges in conveying their research accurately and completely as well as and analyzing it with ease. Based on our findings, authoring environments should consider the context of scientific work aspects such as domain expertise, collaboration culture, and publication traditions. We propose consideration of non-traditional programming languages and environments for scientific work, discuss ways of facilitating interactive visualizations for scientists, and examine ways of meaningfully integrating AI. The critiques, artifacts, and reactions from the scientists along with our analysis and discussion inform how computational tools should be designed for scientific work.

Tue 8 SeptHuman-Computer Interaction
The gist
Scientists often find it hard to show their data clearly and analyze it easily. This paper looks at how creating interactive tools for writing and showing data can help scientists, especially marine biologists. The authors found that tools should fit the scientists’ specific ways of working and sharing results. They also suggest using new kinds of programming and adding artificial intelligence to make these tools better.
Open → 2609.08386v1