arXiv subjects

Haichao Miao

Publications and source records attributed to Haichao Miao.

At least 19 recordsLinked to original sources

arXiv2026

LEDGER: Claim-to-Evidence Trace Graphs for Auditing LLM Agents

Large language model (LLM) agents can now carry out long-horizon technical workflows involving complex tool use, code execution, file edits, and generated artifacts. As agents do more work faster, the productivity bottleneck shifts from producing outputs to auditing whether those outputs are correct and trustworthy. Agent observability systems make fine-grained execution events visible, but visibility alone still leaves reviewers to reconstruct which actions, artifacts, and validation steps matter for a particular conclusion. We introduce LEDGER - Layered Evidence and Decision Graphs for Execution Review, a tracing and review system that builds layered trace graphs over observed agent sessions. LEDGER preserves Trace Records while grouping them into Evidence Nodes and Workflow Nodes, representing artifacts as evidence anchors, and adding typed semantic edges that connect claims to supporting actions, artifacts, and checks. Through data-analysis and coding examples, we show how the resulting traces expose workflow decisions, artifact lineage, repair steps, validation coverage, and claim-support paths for evidence-centered audit.

cs.HC↗

arXiv2026

SciVisAgentBench: A Benchmark for Evaluating Scientific Data Analysis and Visualization Agents

Recent advances in large language models (LLMs) have enabled agentic systems to translate natural-language intent into executable scientific visualization (SciVis) tasks. Despite rapid progress, the community lacks a principled and reproducible benchmark for evaluating these emerging SciVis agents in realistic, multi-step analysis settings. We present SciVisAgentBench, a comprehensive and extensible benchmark for evaluating scientific data analysis and visualization agents. Our benchmark is grounded in a structured taxonomy spanning four dimensions: application domain, data type, complexity level, and visualization operation. It currently comprises 108 expert-crafted cases covering diverse SciVis scenarios. To enable reliable assessment, we introduce a multimodal outcome-centric evaluation pipeline that combines LLM-based judging with deterministic evaluators, including image-based metrics, code checkers, rule-based verifiers, and case-specific evaluators. We also conduct a validity study with 12 SciVis experts to examine the agreement between human and LLM judges. Using this framework, we evaluate representative SciVis agents and general-purpose coding agents to establish initial baselines and reveal capability gaps. SciVisAgentBench is designed as a living benchmark to support systematic comparison, diagnose failure modes, and drive progress in agentic SciVis. The benchmark is available at https://scivisagentbench.github.io/.

cs.AI↗

arXiv2026

A Query-Efficient Stochastic Volume Rendering Framework for Time-Varying Implicit Neural Volumes

Time-varying implicit neural representations (INRs) provide a compact representation of scientific volumes and, for modalities such as dynamic X-ray computed tomography (CT), are often the only practical way to represent the data. However, interactive volume rendering of INRs is challenging, as cheap memory lookups are replaced by expensive neural inferences, hindering the performance. Therefore, conventional volume rendering methods such as ray marching with dense sampling are often impractical. While resampling, caching, and retraining can mitigate this cost, they compromise convenience and accuracy and become impractical for time-varying data. We tackle these challenges using a query-efficient stochastic volume rendering framework based on delta tracking. Our system employs a four-stage pipeline that exploits heterogeneous parallelism, using ray tracing cores for traversal and tensor cores for batched neural evaluation. Furthermore, we present strategies to reduce INR queries via ray budgeting and query pruning, thereby increasing per-frame performance. Using our renderer, many time-varying INRs can be rendered directly from their original representation. The system achieves ~30-40 FPS at 1024x1024 resolution on an RTX 4090 GPU and converges to high-fidelity images. Moreover, the system enables interactive temporal exploration of the continuous domain, with timestep updates taking approximately 1-2 ms.

cs.GR↗

arXiv2026

Exploring LLM Agent Designs and Interaction Modalities for Scientific Visualization

This paper examines how large language model (LLM) agents perform on scientific visualization (SciVis) tasks that require generating visualization workflows from natural-language instructions. We compare three representative agent designs: domain-specific agents with structured tool use, computer-use agents, and general-purpose coding agents, across 15 benchmark tasks, evaluating visualization quality, efficiency, robustness, computational cost, and the impact of persistent memory. We further study interaction modalities, including code scripts, model context protocol (MCP) or API calls, command-line interfaces (CLI), and graphical user interfaces (GUI). Our goal is to characterize the tradeoffs among representative SciVis agent configurations used in practice. The results reveal clear tradeoffs across agent designs and interaction modalities. General-purpose coding agents achieve the highest task success rates but incur greater computational cost, whereas domain-specific agents are more efficient and stable but less flexible. Computer-use agents perform well on individual operations but struggle with multi-step workflows. Across both CLI- and GUI-based settings, persistent memory improves performance over repeated trials, but its effectiveness depends on the interaction mode and the quality of feedback. These findings suggest that future SciVis systems should combine structured tool use, interactive capabilities, and adaptive memory mechanisms to balance performance, robustness, and flexibility.

cs.AI↗

arXiv2026

SciVisAgentSkills: Design and Evaluation of Agent Skills for Scientific Data Analysis and Visualization

Recent advances in agentic visualization have enabled the translation of natural language into executable scientific visualization (SciVis) workflows. While general-purpose coding agents show strong capabilities, they often lack the tool-specific expertise required for SciVis tasks. In this work, we present SciVisAgentSkills, a collection of reusable agent skills that augment coding agents for scientific data analysis and visualization by encoding environment assumptions, tool usage patterns, and domain heuristics across scientific tools such as ParaView, napari, VMD, and TTK. We evaluate these skills on Codex and Claude Code using SciVisAgentBench, a benchmark of 108 expert-designed multi-step tasks. Results show that agent skills improve mean task scores across the evaluated suites, with token-efficiency benefits that depend on the agent harness and tool setting. These findings highlight the importance of structured procedural knowledge for enabling reliable, long-horizon SciVis workflows, while also showing that skills should be studied alongside the execution harness that loads and applies them. The skills are available at https://github.com/KuangshiAi/SciVisAgentSkills.

cs.AI↗

arXiv2026

Toward AI VIS Co-Scientists: A General and End-to-End Agent Harness for Solving Complex Data Visualization Tasks

The ability to inspect, interpret, and communicate complex data is crucial for virtually any scientific endeavor, but often requires significant expertise outside the core domain ranging from data management and analysis to visualization design and implementation. We present an end-to-end agentic harness that, based on only the data and a high level description of the tasks, independently designs custom visual analysis applications (VIS apps). This represents an important step towards a general AI co-scientist envisioned by many as an autonomous system that can autonomously execute long horizon tasks based on high-level directions. Our proposed VIS co-scientist is an essential component of this broader AI co-scientist vision: a harness that can autonomously analyze data and design visualization solutions using a collection of agents and specialized skills that coordinate exploratory analysis, plan, configure the environment, implement, validate the interface, and most importantly evaluate the overall task completion. Each stage produces document and instruction artifacts that guide downstream work and enable iterative refinement. We validate this approach on IEEE SciVis Contests spanning multiple science and engineering fields. These contests serve as ideal proving grounds because they encode real-world complexity: ambiguous requirements, diverse data modalities, design trade-offs, and task-driven validation. Given only the data and target tasks, our system autonomously produces functional single-page VIS Apps with verified linked-view behavior, highly customized to domain experts' specified tasks and needs.

cs.AI↗

arXiv2025

Exploring Bichronous Collaboration in Virtual Environments

Virtual environments (VEs) empower geographically distributed teams to collaborate on a shared project regardless of time. Existing research has separately investigated collaborations within these VEs at the same time (i.e., synchronous) or different times (i.e., asynchronous). In this work, we highlight the often-overlooked concept of bichronous collaboration and define it as the seamless integration of archived information during a real-time collaborative session. We revisit the time-space matrix of computer-supported cooperative work (CSCW) and reclassify the time dimension as a continuum. We describe a system that empowers collaboration across the temporal states of the time continuum within a VE during remote work. We conducted a user study using the system to discover how the bichronous temporal state impacts the user experience during a collaborative inspection. Findings indicate that the bichronous temporal state is beneficial to collaborative activities for information processing, but has drawbacks such as changed interaction and positioning behaviors in the VE.

cs.HC↗

arXiv2025

An Evaluation-Centric Paradigm for Scientific Visualization Agents

Recent advances in multi-modal large language models (MLLMs) have enabled increasingly sophisticated autonomous visualization agents capable of translating user intentions into data visualizations. However, measuring progress and comparing different agents remains challenging, particularly in scientific visualization (SciVis), due to the absence of comprehensive, large-scale benchmarks for evaluating real-world capabilities. This position paper examines the various types of evaluation required for SciVis agents, outlines the associated challenges, provides a simple proof-of-concept evaluation example, and discusses how evaluation benchmarks can facilitate agent self-improvement. We advocate for a broader collaboration to develop a SciVis agentic evaluation benchmark that would not only assess existing capabilities but also drive innovation and stimulate future development in the field.

cs.HC↗

arXiv2025

ParaView-MCP: An Autonomous Visualization Agent with Direct Tool Use

While powerful and well-established, tools like ParaView present a steep learning curve that discourages many potential users. This work introduces ParaView-MCP, an autonomous agent that integrates modern multimodal large language models (MLLMs) with ParaView to not only lower the barrier to entry but also augment ParaView with intelligent decision support. By leveraging the state-of-the-art reasoning, command execution, and vision capabilities of MLLMs, ParaView-MCP enables users to interact with ParaView through natural language and visual inputs. Specifically, our system adopted the Model Context Protocol (MCP) - a standardized interface for model-application communication - that facilitates direct interaction between MLLMs with ParaView's Python API to allow seamless information exchange between the user, the language model, and the visualization tool itself. Furthermore, by implementing a visual feedback mechanism that allows the agent to observe the viewport, we unlock a range of new capabilities, including recreating visualizations from examples, closed-loop visualization parameter updates based on user-defined goals, and even cross-application collaboration involving multiple tools. Broadly, we believe such an agent-driven visualization paradigm can profoundly change the way we interact with visualization tools. We expect a significant uptake in the development of such visualization tools, in both visualization research and industry.

cs.HC↗

arXiv2025

See or Recall: A Sanity Check for the Role of Vision in Solving Visualization Question Answer Tasks with Multimodal LLMs

Recent developments in multimodal large language models (MLLM) have equipped language models to reason about vision and language jointly. This permits MLLMs to both perceive and answer questions about data visualization across a variety of designs and tasks. Applying MLLMs to a broad range of visualization tasks requires us to properly evaluate their capabilities, and the most common way to conduct evaluation is through measuring a model's visualization reasoning capability, analogous to how we would evaluate human understanding of visualizations (e.g., visualization literacy). However, we found that in the context of visualization question answering (VisQA), how an MLLM perceives and reasons about visualizations can be fundamentally different from how humans approach the same problem. During the evaluation, even without visualization, the model could correctly answer a substantial portion of the visualization test questions, regardless of whether any selection options were provided. We hypothesize that the vast amount of knowledge encoded in the language model permits factual recall that supersedes the need to seek information from the visual signal. It raises concerns that the current VisQA evaluation may not fully capture the models' visualization reasoning capabilities. To address this, we propose a comprehensive sanity check framework that integrates a rule-based decision tree and a sanity check table to disentangle the effects of "seeing" (visual processing) and "recall" (reliance on prior knowledge). This validates VisQA datasets for evaluation, highlighting where models are truly "seeing", positively or negatively affected by the factual recall, or relying on inductive biases for question answering. Our study underscores the need for careful consideration in designing future visualization understanding studies when utilizing MLLMs.

cs.HC↗

arXiv2025

Investigating the Influence of Playback Interactivity during Guided Tours for Asynchronous Collaboration in Virtual Reality

Collaborative virtual environments allow workers to contribute to team projects across space and time. While much research has closely examined the problem of working in different spaces at the same time, few have investigated the best practices for collaborating in those spaces at different times aside from textual and auditory annotations. We designed a system that allows experts to record a tour inside a virtual inspection space, preserving knowledge and providing later observers with insights through a 3D playback of the expert's inspection. We also created several interactions to ensure that observers are tracking the tour and remaining engaged. We conducted a user study to evaluate the influence of these interactions on an observing user's information recall and user experience. Findings indicate that independent viewpoint control during a tour enhances the user experience compared to fully passive playback and that additional interactivity can improve auditory and spatial recall of key information conveyed during the tour.

cs.HC↗

arXiv2025

Exploring Multiscale Navigation of Homogeneous and Dense Objects with Progressive Refinement in Virtual Reality

Locating small features in a large, dense object in virtual reality (VR) poses a significant interaction challenge. While existing multiscale techniques support transitions between various levels of scale, they are not focused on handling dense, homogeneous objects with hidden features. We propose a novel approach that applies the concept of progressive refinement to VR navigation, enabling focused inspections. We conducted a user study where we varied two independent variables in our design, navigation style (STRUCTURED vs. UNSTRUCTURED) and display mode (SELECTION vs. EVERYTHING), to better understand their effects on efficiency and awareness during multiscale navigation. Our results showed that unstructured navigation can be faster than structured and that displaying only the selection can be faster than displaying the entire object. However, using an everything display mode can support better location awareness and object understanding.

cs.HC↗

arXiv2024

Bimodal Visualization of Industrial X-Ray and Neutron Computed Tomography Data

Advanced manufacturing creates increasingly complex objects with material compositions that are often difficult to characterize by a single modality. Our collaborating domain scientists are going beyond traditional methods by employing both X-ray and neutron computed tomography to obtain complementary representations expected to better resolve material boundaries. However, the use of two modalities creates its own challenges for visualization, requiring either complex adjustments of bimodal transfer functions or the need for multiple views. Together with experts in nondestructive evaluation, we designed a novel interactive bimodal visualization approach to create a combined view of the co-registered X-ray and neutron acquisitions of industrial objects. Using an automatic topological segmentation of the bivariate histogram of X-ray and neutron values as a starting point, the system provides a simple yet effective interface to easily create, explore, and adjust a bimodal visualization. We propose a widget with simple brushing interactions that enables the user to quickly correct the segmented histogram results. Our semiautomated system enables domain experts to intuitively explore large bimodal datasets without the need for either advanced segmentation algorithms or knowledge of visualization techniques. We demonstrate our approach using synthetic examp

cs.GR↗

arXiv2024

Visualization Literacy of Multimodal Large Language Models: A Comparative Study

The recent introduction of multimodal large language models (MLLMs) combine the inherent power of large language models (LLMs) with the renewed capabilities to reason about the multimodal context. The potential usage scenarios for MLLMs significantly outpace their text-only counterparts. Many recent works in visualization have demonstrated MLLMs' capability to understand and interpret visualization results and explain the content of the visualization to users in natural language. In the machine learning community, the general vision capabilities of MLLMs have been evaluated and tested through various visual understanding benchmarks. However, the ability of MLLMs to accomplish specific visualization tasks based on visual perception has not been properly explored and evaluated, particularly, from a visualization-centric perspective. In this work, we aim to fill the gap by utilizing the concept of visualization literacy to evaluate MLLMs. We assess MLLMs' performance over two popular visualization literacy evaluation datasets (VLAT and mini-VLAT). Under the framework of visualization literacy, we develop a general setup to compare different multimodal large language models (e.g., GPT4-o, Claude 3 Opus, Gemini 1.5 Pro) as well as against existing human baselines. Our study demonstrates MLLMs' competitive performance in visualization literacy, where they outperform humans in certain tasks such as identifying correlations, clusters, and hierarchical structures.

cs.CL↗

arXiv2024

Enabling Additive Manufacturing Part Inspection of Digital Twins via Collaborative Virtual Reality

Digital twins (DTs) are an emerging capability in additive manufacturing (AM), set to revolutionize design optimization, inspection, in situ monitoring, and root cause analysis. AM DTs typically incorporate multimodal data streams, ranging from machine toolpaths and in-process imaging to X-ray CT scans and performance metrics. Despite the evolution of DT platforms, challenges remain in effectively inspecting them for actionable insights, either individually or in a multidisciplinary team setting. Quality assurance, manufacturing departments, pilot labs, and plant operations must collaborate closely to reliably produce parts at scale. This is particularly crucial in AM where complex structures require a collaborative and multidisciplinary approach. Additionally, the large-scale data originating from different modalities and their inherent 3D nature pose significant hurdles for traditional 2D desktop-based inspection methods. To address these challenges and increase the value proposition of DTs, we introduce a novel virtual reality (VR) framework to facilitate collaborative and real-time inspection of DTs in AM. This framework includes advanced features for intuitive alignment and visualization of multimodal data, visual occlusion management, streaming large-scale volumetric data, and collaborative tools, substantially improving the inspection of AM components and processes to fully exploit the potential of DTs in AM.

cs.HC↗

arXiv2024

A Virtual Environment for Collaborative Inspection in Additive Manufacturing

Additive manufacturing (AM) techniques have been used to enhance the design and fabrication of complex components for various applications in the medical, aerospace, energy, and consumer products industries. A defining feature for many AM parts is the complex internal geometry enabled by the printing process. However, inspecting these internal structures requires volumetric imaging, i.e., X-ray CT, leading to the well-known challenge of visualizing complex 3D geometries using 2D desktop interfaces. Furthermore, existing tools are limited to single-user systems making it difficult to jointly discuss or share findings with a larger team, i.e., the designers, manufacturing experts, and evaluation team. In this work, we present a collaborative virtual reality (VR) for the exploration and inspection of AM parts. Geographically separated experts can virtually inspect and jointly discuss data. It also supports VR and non-VR users, who can be spectators in the VR environment. Various features for data exploration and inspection are developed and enhanced via real-time synchronization. We followed usability and interface verification guidelines using Nielsen's heuristics approach. Furthermore, we conducted exploratory and semi-structured interviews with domain experts to collect qualitative feedback. Results reveal potential benefits, applicability, and current limitations. The proposed collaborative VR environment provides a new basis and opens new research directions for virtual inspection and team collaboration in AM settings.

cs.HC↗

arXiv2023

AVA: Towards Autonomous Visualization Agents through Visual Perception-Driven Decision-Making

With recent advances in multi-modal foundation models, the previously text-only large language models (LLM) have evolved to incorporate visual input, opening up unprecedented opportunities for various applications in visualization. Our work explores the utilization of the visual perception ability of multi-modal LLMs to develop Autonomous Visualization Agents (AVAs) that can interpret and accomplish user-defined visualization objectives through natural language. We propose the first framework for the design of AVAs and present several usage scenarios intended to demonstrate the general applicability of the proposed paradigm. The addition of visual perception allows AVAs to act as the virtual visualization assistant for domain experts who may lack the knowledge or expertise in fine-tuning visualization outputs. Our preliminary exploration and proof-of-concept agents suggest that this approach can be widely applicable whenever the choices of appropriate visualization parameters require the interpretation of previous visual output. Feedback from unstructured interviews with experts in AI research, medical visualization, and radiology has been incorporated, highlighting the practicality and potential of AVAs. Our study indicates that AVAs represent a general paradigm for designing intelligent visualization systems that can achieve high-level visualization goals, which pave the way for developing expert-level visualization agents in the future.

cs.HC↗

arXiv2022

SynopSet: Multiscale Visual Abstraction Set for Explanatory Analysis of DNA Nanotechnology Simulations

We propose a new abstraction set (SynopSet) that has a continuum of visual representations for the explanatory analysis of molecular dynamics simulations (MDS) in the DNA nanotechnology domain. By re-purposing the commonly used progress bar and designing novel visuals, as well as transforming the data from the domain format to a format that better fits the newly designed visuals, we compose this new set of representations. This set is also designed to be capable of showing all spatial and temporal details, and all structural complexity, or abstracting these to various degrees, enabling both the slow playback of the simulation for detailed examinations or very fast playback for an overview that helps to efficiently identify events of interest, as well as several intermediate levels between these two extremes. For any pair of successive representations, we demonstrate smooth, continuous transitions, enabling users to keep track of relevant information from one representation to the next. By providing multiple representations suited to different temporal resolutions and connected by smooth transitions, we enable time-efficient simulation analysis, giving users the opportunity to examine and present important phases in great detail, or leverage abstract representations to go over uneventful phases much faster. Domain experts can thus gain actionable insight about their simulations and communicate it in a much shorter time. Further, the novel representations are more intuitive and also enable researchers unfamiliar with MDS analysis graphs to better understand the simulation results. We assessed the effectiveness of SynopSet on 12 DNA nanostructure simulations together with a domain expert. We have also shown that our set of representations can be systematically located in a visualization space, dubbed SynopSpace.

q-bio.QM↗