Large language model (LLM)-based multi-agent systems have demonstrated impressive capabilities in handling complex tasks. However, the complexity of agentic behaviors makes these systems difficult to understand. When failures occur, developers often struggle to identify root causes and to determine actionable paths for improvement. Traditional methods that rely on inspecting raw log records are inefficient, given both the large volume and complexity of data. To address this challenge, we propose a framework and an interactive system, DiLLS, designed to reveal and structure the behaviors of multi-agent systems. The key idea is to organize information across three levels of query completion: activities, actions, and operations. By probing the multi-agent system through natural language, DiLLS derives and organizes information about planning and execution into a structured, multi-layered summary. Through a user study, we show that DiLLS significantly improves developers’ effectiveness and efficiency in identifying, diagnosing, and understanding failures in LLM-based multi-agent systems.

Understanding the behavior of large language models (LLMs) is crucial for ensuring their safe and reliable use. However, existing explainable AI (XAI) methods for LLMs primarily rely on word-level explanations, which are often computationally inefficient and misaligned with human reasoning processes. Moreover, these methods often treat explanation as a one-time output, overlooking its inherently interactive and iterative nature. In this paper, we present LLM Analyzer, an interactive visualization system that addresses these limitations by enabling intuitive and efficient exploration of LLM behaviors through counterfactual analysis. Our system features a novel algorithm that generates fluent and semantically meaningful counterfactuals via targeted removal and replacement operations at user-defined levels of granularity. These counterfactuals are used to compute feature attribution scores, which are then integrated with concrete examples in a table-based visualization, supporting dynamic analysis of model behavior. A user study with LLM practitioners and interviews with experts demonstrate the system's usability and effectiveness, emphasizing the importance of involving humans in the explanation process as active participants rather than passive recipients.

Human preferences are widely used to align large language models (LLMs) through methods such as reinforcement learning from human feedback (RLHF).However, the current user interfaces require annotators to compare text paragraphs, which is cognitively challenging when the texts are long or unfamiliar.This paper contributes by studying the decomposition principle as an approach to improving the quality of human feedback for LLM alignment.This approach breaks down the text into individual claims instead of directly comparing two long-form text responses.Based on the principle, we build a novel user interface DxHF.It enhances the comparison process by showing decomposed claims, visually encoding the relevance of claims to the conversation and linking similar claims.This allows users to skim through key information and identify differences for better and quicker judgment.Our technical evaluation shows evidence that decomposition generally improves feedback accuracy regarding the ground truth, particularly for users with uncertainty.A crowdsourcing study with 160 participants indicates that using DxHF improves feedback accuracy by an average of 5%, although it increases the average feedback time by 18 seconds.Notably, accuracy is significantly higher in situations where users have less certainty.The finding of the study highlights the potential of HCI as an effective method for improving human-AI alignment.

Large Language Models (LLMs) are notorious for blending fact with fiction and generating non-factual content, known as hallucinations. To address this challenge, we propose an interactive system that helps users gain insight into the reliability of the generated text. Our approach is based on the idea that the self-consistency of multiple samples generated by the same LLM relates to its confidence in individual claims in the generated texts. Using this idea, we design RELIC, an interactive system that enables users to investigate and verify semantic-level variations in multiple long-form responses. This allows users to recognize potentially inaccurate information in the generated text and make necessary corrections. From a user study with ten participants, we demonstrate that our approach helps users better verify the reliability of the generated text. We further summarize the design implications and lessons learned from this research for future studies of reliable human-LLM interactions.

Benchmark datasets play an important role in evaluating Natural Language Understanding (NLU) models. However, shortcuts-unwanted biases in the benchmark datasets-can damage the effectiveness of benchmark datasets in revealing models' real capabilities. Since shortcuts vary in coverage, productivity, and semantic meaning, it is challenging for NLU experts to systematically understand and avoid them when creating benchmark datasets. In this paper, we develop a visual analytics system, ShortcutLens, to help NLU experts explore shortcuts in NLU benchmark datasets. The system allows users to conduct multi-level exploration of shortcuts. Specifically, Statistics View helps users grasp the statistics such as coverage and productivity of shortcuts in the benchmark dataset. Template View employs hierarchical and interpretable templates to summarize different types of shortcuts. Instance View allows users to check the corresponding instances covered by the shortcuts. We conduct case studies and expert interviews to evaluate the effectiveness and usability of the system. The results demonstrate that ShortcutLens supports users in gaining a better understanding of benchmark dataset issues through shortcuts, inspiring them to create challenging and pertinent benchmark datasets.

Simulation-based Medical Education (SBME) has been developed as a cost-effective means of enhancing the diagnostic skills of novice physicians and interns, thereby mitigating the need for resource-intensive mentor-apprentice training. However, feedback provided in most SBME is often directed towards improving the operational proficiency of learners, rather than providing summative medical diagnoses that result from experience and time. Additionally, the multimodal nature of medical data during diagnosis poses significant challenges for interns and novice physicians, including the tendency to overlook or over-rely on data from certain modalities, and difficulties in comprehending potential associations between modalities. To address these challenges, we present DiagnosisAssistant, a visual analytics system that leverages historical medical records as a proxy for multimodal modeling and visualization to enhance the learning experience of interns and novice physicians. The system employs elaborately designed visualizations to explore different modality data, offer diagnostic interpretive hints based on the constructed model, and enable comparative analyses of specific patients. Our approach is validated through two case studies and expert interviews, demonstrating its effectiveness in enhancing medical training.

Non-fungible tokens (NFTs) have recently gained widespread popularity as an alternative investment. However, the lack of assessment criteria has caused intense volatility in NFT marketplaces. Identifying attributes impacting the market performance of NFT collectibles is crucial but challenging due to the massive amount of heterogeneous and multi-modal data in NFT transactions, e.g., social media texts, numerical trading data, and images. To address this challenge, we introduce an interactive dual-centric visual analytics system, NFTeller, to facilitate users’ analysis. First, we collaborate with five domain experts to distill static and dynamic impact attributes and collect relevant data. Next, we derive six analysis tasks and develop NFTeller to present the evolution of NFT transactions and correlate NFTs’ market performance with impact attributes. Notably, we create an augmented chord diagram with a radial stacked bar chart to explore intersections between NFT collection projects and whale accounts. Finally, we conduct three case studies and interview domain experts to evaluate the effectiveness and usability of this system. As such, we gain in-depth insights into assessing NFT collectibles and detecting opportune moments for investment.

Reference-based cell-type annotation can significantly reduce time and effort in single-cell analysis by transferring labels from a previously-annotated dataset to a new dataset. However, label transfer by end-to-end computational methods is challenging due to the entanglement of technical (e.g., from different sequencing batches or techniques) and biological (e.g., from different cellular microenvironments) variations, only the first of which must be removed. To address this issue, we propose Polyphony, an interactive transfer learning (ITL) framework, to complement biologists' knowledge with advanced computational methods. Polyphony is motivated and guided by domain experts' needs for a controllable, interactive, and algorithm-assisted annotation process, identified through interviews with seven biologists. We introduce anchors, i.e., analogous cell populations across datasets, as a paradigm to explain the computational process and collect user feedback for model improvement. We further design a set of visualizations and interactions to empower users to add, delete, or modify anchors, resulting in refined cell type annotations. The effectiveness of this approach is demonstrated through quantitative experiments, two hypothetical use cases, and interviews with two biologists. The results show that our anchor-based ITL method takes advantage of both human and machine intelligence in annotating massive single-cell datasets.

The last decade has witnessed many visual analytics (VA) systems that make successful applications to wide-ranging domains like urban analytics and explainable AI. However, their research rigor and contributions have been extensively challenged within the visualization community. We come in defence of VA systems by contributing two interview studies for gathering critics and responses to those criticisms. First, we interview 24 researchers to collect criticisms the review comments on their VA work. Through an iterative coding and refinement process, the interview feedback is summarized into a list of 36 common criticisms. Second, we interview 17 researchers to validate our list and collect their responses, thereby discussing implications for defending and improving the scientific values and rigor of VA systems. We highlight that the presented knowledge is deep, extensive, but also imperfect, provocative, and controversial, and thus recommend reading with an inclusive and critical eye. We hope our work can provide thoughts and foundations for conducting VA research and spark discussions to promote the research field forward more rigorously and vibrantly.

Machine learning (ML) is increasingly applied to Electronic Health Records (EHRs) to solve clinical prediction tasks. Although many ML models perform promisingly, issues with model transparency and interpretability limit their adoption in clinical practice. Directly using existing explainable ML techniques in clinical settings can be challenging. Through literature surveys and collaborations with six clinicians with an average of 17 years of clinical experience, we identified three key challenges, including clinicians' unfamiliarity with ML features, lack of contextual information, and the need for cohort-level evidence. Following an iterative design process, we further designed and developed VBridge, a visual analytics tool that seamlessly incorporates ML explanations into clinicians' decision-making workflow. The system includes a novel hierarchical display of contribution-based feature explanations and enriched interactions that connect the dots between ML features, explanations, and data. We demonstrated the effectiveness of VBridge through two case studies and expert interviews with four clinicians, showing that visually associating model explanations with patients' situational records can help clinicians better interpret and use model predictions when making clinician decisions. We further derived a list of design implications for developing future explainable ML tools to support clinical decision-making.

With machine learning models being increasingly applied to various decision-making scenarios, people have spent growing efforts to make machine learning models more transparent and explainable. Among various explanation techniques, counterfactual explanations have the advantages of being human-friendly and actionable-a counterfactual explanation tells the user how to gain the desired prediction with minimal changes to the input. Besides, counterfactual explanations can also serve as efficient probes to the models' decisions. In this work, we exploit the potential of counterfactual explanations to understand and explore the behavior of machine learning models. We design DECE, an interactive visualization system that helps understand and explore a model's decisions on individual instances and data subsets, supporting users ranging from decision-subjects to model developers. DECE supports exploratory analysis of model decisions by combining the strengths of counterfactual explanations at instance- and subgroup-levels. We also introduce a set of interactions that enable users to customize the generation of counterfactual explanations to find more actionable ones that can suit their needs. Through three use cases and an expert interview, we demonstrate the effectiveness of DECE in supporting decision exploration tasks and instance explanations.

Recently we have witnessed growing adoption of deep sequence models (e.g. LSTMs) in many application domains, including predictive health care, natural language processing, and log analysis. However, the intricate working mechanism of these models confines their accessibility to the domain experts. Their black-box nature also makes it a challenging task to incorporate domain-specific knowledge of the experts into the model. In ProtoSteer (Prototype Steering), we tackle the challenge of directly involving the domain experts to steer a deep sequence model without relying on model developers as intermediaries. Our approach originates in case-based reasoning, which imitates the common human problem-solving process of consulting past experiences to solve new problems. We utilize ProSeNet (Prototype Sequence Network), which learns a small set of exemplar cases (i.e., prototypes) from historical data. In ProtoSteer they serve both as an efficient visual summary of the original data and explanations of model decisions. With ProtoSteer the domain experts can inspect, critique, and revise the prototypes interactively. The system then incorporates user-specified prototypes and incrementally updates the model. We conduct extensive case studies and expert interviews in application domains including sentiment analysis on texts and predictive diagnostics based on vehicle fault logs. The results demonstrate that involvements of domain users can help obtain more interpretable models with concise prototypes while retaining similar accuracy.