Fully autonomous teams of LLM-powered AI agents are emerging that collaborate to perform complex tasks for users. What challenges do developers face when trying to build and debug these AI agent teams? In formative interviews with five AI agent developers, we identify core challenges: difficulty reviewing long agent conversations to localize errors, lack of support in current tools for interactive debugging, and the need for tool support to iterate on agent configuration. Based on these needs, we developed an interactive multi-agent debugging tool, AGDebugger, with a UI for browsing and sending messages, the ability to edit and reset prior agent messages, and an overview visualization for navigating complex message histories. In a two-part user study with 14 participants, we identify common user strategies for steering agents and highlight the importance of interactive message resets for debugging. Our studies deepen understanding of interfaces for debugging increasingly important agentic workflows.

Statistical practices such as building regression models or running hypothesis tests rely on following rigorous procedures of steps and verifying assumptions on data to produce valid results. However, common statistical tools do not verify users’ decision choices and provide low-level statistical functions without instructions on the whole analysis practice. Users can easily misuse analysis methods, potentially decreasing the validity of results. To address this problem, we introduce GuidedStats, an interactive interface within computational notebooks that encapsulates guidance, models, visualization, and exportable results into interactive workflows. It breaks down typical analysis processes, such as linear regression and two-sample T-tests, into interactive steps supplemented with automatic visualizations and explanations for step-wise evaluation. Users can iterate on input choices to refine their models, while recommended actions and exports allow the user to continue their analysis in code. Case studies show how GuidedStats offers valuable instructions for conducting fluid statistical analyses while finding possible assumption violations in the underlying data, supporting flexible and accurate statistical analyses.

Profiling data by plotting distributions and analyzing summary statistics is a critical step throughout data analysis. Currently, this process is manual and tedious since analysts must write extra code to examine their data after every transformation. This inefficiency may lead to data scientists profiling their data infrequently, rather than after each transformation, making it easy for them to miss important errors or insights. We propose continuous data profiling as a process that allows analysts to immediately see interactive visual summaries of their data throughout their data analysis to facilitate fast and thorough analysis. Our system, AutoProfiler, presents three ways to support continuous data profiling: (1) it automatically displays data distributions and summary statistics to facilitate data comprehension; (2) it is live, so visualizations are always accessible and update automatically as the data updates; (3) it supports follow up analysis and documentation by authoring code for the user in the notebook. In a user study with 16 participants, we evaluate two versions of our system that integrate different levels of automation: both automatically show data profiles and facilitate code authoring, however, one version updates reactively ("live") and the other updates only on demand ("dead"). We find that both tools, dead or alive, facilitate insight discovery with 91% of user-generated insights originating from the tools rather than manual profiling code written by users. Participants found live updates intuitive and felt it helped them verify their transformations while those with on-demand profiles liked the ability to look at past visualizations. We also present a longitudinal case study on how AutoProfiler helped domain scientists find serendipitous insights about their data through automatic, live data profiles. Our results have implications for the design of future tools that offer automated data analysis support.

Data science is characterized by evolution: since data science is exploratory, results evolve from moment to moment; since it can be collaborative, results evolve as the work changes hands. While existing tools help data scientists track changes in code, they provide less support for understanding the iterative changes that the code produces in the data. We explore the idea of visualizing differences in datasets as a core feature of exploratory data analysis, a concept we call Diff in the Loop (DITL). We evaluated DITL in a user study with 16 professional data scientists and found it helped them understand the implications of their actions when manipulating data. We summarize these findings and discuss how the approach can be generalized to different data science workflows.

Abstract Existing visualization recommendation systems commonly rely on a single snapshot of a dataset to suggest visualizations to users. However, exploratory data analysis involves a series of related interactions with a dataset over time rather than one‐off analytical steps. We present Solas, a tool that tracks the history of a user's data analysis, models their interest in each column, and uses this information to provide visualization recommendations, all within the user's native analytical environment. Recommending with analysis history improves visualizations in three primary ways: task‐specific visualizations use the provenance of data to provide sensible encodings for common analysis functions, aggregated history is used to rank visualizations by our model of a user's interest in each column, and column data types are inferred based on applied operations. We present a usage scenario and a user evaluation demonstrating how leveraging analysis history improves in situ visualization recommendations on real‐world analysis tasks.

With the widespread use of toxic language online, platforms are increasingly using automated systems that leverage advances in natural language processing to automatically flag and remove toxic comments. However, most automated systems, when detecting and moderating toxic language, do not provide feedback to their users, let alone provide an avenue of recourse for these users to make actionable changes. We present our work, RECAST, an interactive, open-sourced web tool for visualizing these models' toxic predictions, while providing alternative suggestions for flagged toxic language. Our work also provides users with a new path of recourse when using these automated moderation tools. RECAST highlights text responsible for classifying toxicity, and allows users to interactively substitute potentially toxic phrases with neutral alternatives. We examined the effect of RECAST via two large-scale user evaluations, and found that RECAST was highly effective at helping users reduce toxicity as detected through the model. Users also gained a stronger understanding of the underlying toxicity criterion used by black-box models, enabling transparency and recourse. In addition, we found that when users focus on optimizing language for these models instead of their own judgement (which is the implied incentive and goal of deploying automated models), these models cease to be effective classifiers of toxicity compared to human annotations. This opens a discussion for how toxicity detection models work and should work, and their effect on the future of online discourse.