Caught by Surprise, Caught by Culture: Bridging Facial Expression's Recognition and Interpretation of Surprise Across CulturesFacial expressions are powerful signals of human emotion, shaping both human–human and human–computer interaction. As interactive technologies, from adaptive interfaces to emotion-aware agents, become more pervasive, systems are increasingly expected to recognize and respond to users' emotions naturally. But what if a system misreads your face? Such misinterpretation is particularly likely when cultural differences in emotion perception are overlooked. This problem may be compounded by the fact that most facial emotion recognition (FER) models are trained on datasets that reflect the norms of a particular cultural group that assume universality, limiting their reliability in multicultural contexts. Surprise, in particular, is an emotion whose valence can be either positive or negative depending on context, making it a critical case for investigating cultural bias in FER. To address this, we examined how cultural background shapes the recognition and valence interpretation of surprise facial expressions among South Korean (N=36) and American (N=34) participants. Participants labeled 200 facial expressions (surprise and fear), rated their perceived valence, and described personal experiences of surprise. Results show that South Korean-labeled surprise expressions exhibited stronger negative Action Unit (AU) activation and lower valence ratings, whereas American-labeled ones showed more balanced or positive facial cues. Qualitative accounts further revealed that South Koreans framed surprise as tense or socially cautious, while Americans viewed it as open and situationally flexible. These findings bridge recognition and interpretation in cross-cultural emotion research and highlight the need for culturally adaptive FER systems that can interpret ambiguous emotions like surprise more inclusively.2026HJHa Eun Jang et al.Yonsei UniversityEmotion Recognition & DetectionAffective Feedback & Emotion Regulation InterfacesMultilingual & Cross-Cultural Voice InteractionIUI
ChoiceMates: Supporting Unfamiliar Online Decision-Making with Multi-Agent Conversational InteractionsFrom purchasing a gift to deciding on a hobby, unfamiliar decisions---decisions without domain knowledge and experience---are frequent and significant. The complexity and uncertainty of such decisions demand unique approaches to information seeking, understanding, and decision-making. Our formative study highlights that in the current workflow, users want to start by discovering broad and relevant domain information evenly and simultaneously, quickly address emerging inquiries, and gain personalized standards to assess information found. We present ChoiceMates, an interactive multi-agent system designed to address these needs by enabling users to engage with a dynamic set of LLM agents each presenting a unique experience in the domain. Unlike existing multi-agent systems that automate tasks with agents, the user orchestrates agents to assist their decision-making process in each turn, through chatting with all agents, with a tagged subset of agents, or calling in new agents into the space. By comparing ChoiceMates with a web search condition and a multi-agent framework (n=12), we show that ChoiceMates enables a more confident, satisfactory decision-making with better situation understanding than web search, and higher decision quality than a commercial multi-agent framework. We further illustrate how participants utilized ChoiceMates to make unfamiliar decisions, providing insights into designing a more controllable and collaborative multi-agent system.2026JPJeongeon Park et al.University of California San DiegoHuman-LLM CollaborationAI-Assisted Decision-Making & AutomationConversational ChatbotsIUI
Sustainable Human-AI Collaboration in Creative Contexts: An Integrated Approach of CASA, UTAUT, Psychological Ownership, and Self-Determination TheoryThis study examines the psychological mechanisms that enable sustainable human–AI collaboration in creative tasks. Drawing on the CASA paradigm, we frame generative AI as a social collaborator and integrate UTAUT, psychological ownership theory, and self-determination theory to explain users’ continued engagement with AI tools. We test how AI performance expectancy influences psychological ownership, collaboration satisfaction, and continuance intention, and whether these mechanisms vary by creative context (pure vs. work-related) or collaboration type (human-led vs. AI-led). Results show that performance expectancy enhances ownership, satisfaction, and continuance intention, with ownership and satisfaction further reinforcing continued use. However, in AI-led collaboration, its positive effect on satisfaction is weakened, while creative context shows no significant differences, suggesting that core psychological processes generalize across creative purposes. This study extends UTAUT by incorporating psychological mechanisms into human–AI collaboration and provides a theoretical basis for sustainable use of generative AI.2026SHSemin Hong et al.Yonsei UniversityGenerative AI (Text, Image, Music, Video)Human-LLM CollaborationAI-Assisted Decision-Making & AutomationCHI
Understanding Spatiotemporal-Aware Multimodal Conversational Search in the Outdoor Urban SpaceEmerging multimodal conversational search (MCS) tools (e.g., Gemini Live) allow users to search for spatiotemporal information through natural language dialogues as they move through urban space. Despite the growing popularity of these tools, there is limited understanding of how people engage with this technology. To address this gap, we developed UrbanSearch, an MCS technology probe designed to capture the user's current geolocation, time, and visual surroundings. A contextual inquiry (N=23) revealed that MCS tools provide two core values: requiring low effort in forming queries while offering highly relevant responses, and functioning as a central information gateway. As a promising technology, MCS supports environmental learning, in-situ decision making, and personalized navigation. Participants also revealed unmet needs for spatial reasoning and transparent integration of multi-source information, along with concerns related to peripheral awareness, social context, and personal space. Drawing from the findings, we discuss design implications for future MCS tools in urban spaces.2026JXJiangnan Xu et al.Rochester Institute of TechnologyExploratory Search & Information SeekingConversational Search & QA SystemsContext-Aware ComputingCHI
Data-Prompt Co-Evolution: Growing Test Sets to Refine LLM BehaviorLarge Language Models (LLMs) are increasingly embedded in applications, and people can shape model behavior by editing prompt instructions. Yet encoding subtle, domain-specific policies into prompts is challenging. Although this process often benefits from concrete test cases, test data and prompt instructions are typically developed as separate artifacts, reflecting traditional machine learning practices in which model tuning was slow and test sets were static. We argue that the fast, iterative nature of prompt engineering calls for removing this separation and enabling a new workflow: data-prompt co-evolution, where a living test set and prompt instructions evolve in tandem. We present an interactive system that operationalizes this workflow. It guides application developers to discover edge cases, articulate rationales for desired behavior, and iteratively evaluate revised prompts against a growing test set. A user study shows our workflow helps people refine prompts systematically, better aligning them with their intended policies. This work points toward more robust and responsible LLM applications through human-in-the-loop development.2026MLMinjae Lee et al.Yonsei UniversityHuman-LLM CollaborationAI-Assisted Decision-Making & AutomationUser Research Methods (Interviews, Surveys, Observation)CHI
"Do I Really Need This?": Illuminating Challenges in Integrating Computational Training Tools in Esports CoachingThe rise in popularity and value of esports motivates the creation of computational training tools (CTTs) for learning, assessment, and skill gain. While some tools exist commercially, much of the work in the research literature is rarely used outside of a lab, resulting in a lack of knowledge on the challenges involved in real-world integration. In this work, we develop a bespoke CTT for League of Legends, MySkills, based on prior work and deploy it at a professional training academy for three months. Based on two rounds of stakeholder interviews, we uncover insights into users' perspectives on using CTTs in esports coaching and the challenges inherent in introducing a novel tool into an existing, real-world esports training context. From these results, we connect the domain of esports training technology to existing conversations on translational HCI, challenges in bridging research and practice, and present implications for future work.2026EKErica Kleinman et al.Northeastern UniversityGame UX & Player BehaviorSerious & Functional GamesPrototyping & User TestingCHI
Seen but Ignored: Understanding User Disengagement from Emergency Alerts in High-Frequency Contexts — A Case Study of South KoreaPublic Warning Systems (PWS) are critical infrastructures for protecting lives during emergencies, yet many users increasingly ignore or disable alerts. Prior research has focused on attentive recipients, overlooking those who disengage mentally or behaviorally. We examine disengagement as a gradual process of psychological detachment shaped by alert fatigue, trust erosion, and perceived inefficacy. Focusing on South Korea’s high-frequency cell broadcast system, averaging 80 messages per day, we conducted a qualitative study with 37 participants classified as responders, ignorers, or blockers, drawing on EPPM and PADM. Through interactive message evaluation and interviews, we traced cognitive and emotional pathways from message reception to protective action or inaction. Our findings reveal structural and psychological barriers, including fixed cognitive anchors that preemptively dismiss alerts, information-seeking behaviors rarely leading to action, and divergent adaptations to repeated false alarms. We reframe emergency alerts as adaptive user–system interfaces shaped by cumulative experience, not static channels. We show how PADM pathways become non-linear, truncated, or collapsed under saturated alert environments. We contribute design implications for more adaptive, trustworthy, and user-sensitive emergency alert systems.2026JHJuhye Ha et al.Graduate School of Information Yonsei UniversityEmergency Communication & Early Warning SystemsHCI in Public Health Crises (e.g., COVID-19)Privacy Perception & Decision-MakingCHI
The Timing of Breaks for Resilience: Collective Recovery in Multi-User Virtual RealityThe pursuit of seamlessness in collaborative VR often creates a paradox: concealing technical failures generates asymmetric awareness, fracturing the shared reality essential for teamwork. We argue instead that disruption timing acts as an information structure. Drawing on the theory of rational rituals, we posit that a simultaneous onset creates a Public, Synchronous, Bounded (PSB) anchor that establishes common knowledge. We tested this framework with 34 triads (N = 102) performing interdependent tasks. Results show that simultaneous disruptions significantly accelerated Time-to-Recovery (TTR) and preserved role stability by enabling a compact A–R–E sequence (affect-check, reorientation, re-entry). Conversely, asynchronous onsets caused epistemic fragmentation and role churn. We contribute the coordination wrapper, a design strategy that transforms inevitable system failures into synthetic PSB cues, shifting the paradigm from error minimization to resilient recovery.2026HKHayeon Kim et al.Dankook UniversitySocial & Collaborative VRImmersion & Presence ResearchPrototyping & User TestingCHI
DanXeReflect: Interacting with the Spatio-Temporal Past Movements for Embodied, Reflective Choreographic CollaborationChoreographic reflection relies on iterative dialogue, where dancers and choreographers refine movement through embodied demonstration and shared feedback in studio rehearsal. With the shift to video, this exchange becomes constrained: annotations detach from the body, gestures lose spatial grounding, and subtle variations are difficult to capture. Advances in markerless motion capture enable 3D reconstruction from rehearsal video, allowing past recordings to be re-materialized for embodied interaction in XR. We present DanXeReflect, an XR system that transforms flat video into a virtual studio where movements appear as interactive avatars. Users can re-enact poses to search sequences, perform alternative revisions alongside originals, and attach annotations directly to body parts. A study with choreographers and dancers shows how these embodied interactions reposition spatio-temporal data as collaborative anchors, extending reflective dialogue beyond co-located rehearsal into asynchronous, distributed practice.2026HKHyunju Kim et al.Cornell UniversitySocial & Collaborative VRIdentity & Avatars in XRFull-Body Interaction & Embodied InputCHI
Cerebra: Aligning Implicit Knowledge in Interactive SQL AuthoringLLM-driven tools have significantly lowered barriers to writing SQL queries. However, user instructions are often underspecified, assuming the model understands implicit knowledge, such as dataset schemas, domain conventions, and task-specific requirements, that isn't explicitly provided. This results in frequently erroneous scripts that require users to repeatedly clarify their intent. Additionally, users struggle to validate generated scripts because they cannot verify whether the model correctly applied implicit knowledge. We present Cerebra, an interactive NL-to-SQL tool that aligns implicit knowledge between users and LLMs during SQL authoring. Cerebra automatically retrieves implicit knowledge from historical SQL scripts based on user instructions, presents this knowledge in an interactive tree view for code review, and supports iterative refinement to improve generated scripts. To evaluate the effectiveness and usability of Cerebra, we conducted a user study with 16 participants, demonstrating its improved support for customized SQL authoring. The source code of Cerebra is available at https://github.com/zjuidg/CHI26-Cerebra.2026YZYunfan Zhou et al.Zhejiang UniversityHuman-LLM CollaborationExplainable AI (XAI)Computational Methods in HCICHI
Simulating Human Audiovisual Search BehaviorLocating a target based on auditory and visual cues—such as finding a car in a crowded parking lot or identifying a speaker in a virtual meeting—requires balancing effort, time, and accuracy under uncertainty. Existing models of audiovisual search often treat perception and action in isolation, overlooking how people adaptively coordinate movement and sensory strategies. We present Sensonaut, a computational model of embodied audiovisual search. The core assumption is that people deploy their body and sensory systems in ways they believe will most efficiently improve their chances of locating a target, trading off time and effort under perceptual constraints. Our model formulates this as a resource-rational decision-making problem under partial observability. We validate the model against newly collected human data, showing that it reproduces both adaptive scaling of search time and effort under task complexity, occlusion, and distraction, and characteristic human errors. Our simulation of human-like resource-rational search informs the design of audiovisual interfaces that minimize search cost and cognitive load.2026HCHyunsung Cho et al.Aalto UniversityEye Tracking & Gaze InteractionSonification & Auditory DisplayAffective Feedback & Emotion Regulation InterfacesCHI
HyPockeTuner: Bringing Hyperparameter Optimization to Mobile DevicesHyperparameter optimization (HPO) is a long-running process that can span hours or even days. While recent Human-in-the-Loop HPO systems enable monitoring and steering of the process, they are typically designed for desktop environments, which limits their effectiveness in managing prolonged experiments in practice. To address these limitations, we present HyPockeTuner, an interactive mobile system that enables users to monitor, steer, and reflect on HPO experiments anytime, anywhere from smartphones. Its mobile-tailored interface supports tracking experiment history and visualizing the relationship between user interventions and performance changes. HyPockeTuner also employs a notification workflow that alerts users to important events, reducing the burden of constant monitoring while enabling timely interventions. In a pilot study, we validated that users could readily identify critical events, such as performance improvements and intervention points, through our visualization. Furthermore, two five-day deployment studies with follow-up reflection sessions demonstrated that users could integrate experiment management into their daily routines and reflect on past decisions, generating insights for future improvement.2026DHDonghee Hong et al.Sungkyunkwan UniversityAutoML InterfacesRemote Work Tools & ExperienceBehavior Change & Reflection TechnologyCHI
LingoQ: Bridging the Gap between EFL Learning and Work through AI-Generated Work-Related QuizzesNon-native English speakers performing English-related tasks at work struggle to sustain EFL learning, despite their motivation. Often, study materials are disconnected from their work context. Our formative study revealed that reviewing work-related English becomes burdensome with current systems, especially after work. Although workers rely on LLM-based assistants to address their immediate needs, these interactions may not directly contribute to their English skills. We present LingoQ, an AI-mediated system that allows workers to practice English using quizzes generated from their LLM queries during work. LingoQ leverages these on-the-fly queries using AI to generate personalized quizzes that workers can review and practice on their smartphones. We conducted a three-week deployment study with 28 EFL workers to evaluate LingoQ. Participants valued the quality-assured, work-situated quizzes and constantly engaging with the app during the study. This active engagement improved self-efficacy and led to learning gains for beginners and, potentially, for intermediate learners. Drawing on these results, we discuss design implications for leveraging workers' growing reliance on LLMs to foster proficiency and engagement while respecting work boundaries and ethics.2026YYYeonsun Yang et al.DGISTHuman-LLM CollaborationProgramming Education & Computational ThinkingIntelligent Tutoring Systems & Learning AnalyticsCHI
State Your Intention to Steer Your Attention: An AI Assistant for Intentional Digital LivingWhen working on digital devices, people often face distractions that can lead to a decline in productivity and efficiency, as well as negative psychological and emotional impacts. To address this challenge, we introduce a novel Artificial Intelligence (AI) assistant that elicits a user's intention, assesses whether ongoing activities are in line with that intention, and provides gentle nudges when deviations occur. The system leverages a large language model to analyze screenshots, application titles, and URLs, issuing notifications when behavior diverges from the stated goal. Its detection accuracy is refined through initial clarification dialogues and continuous user feedback. In a three-week, within-subjects field deployment with 22 participants, we compared our assistant to both a rule-based intent reminder system and a passive baseline that only logged activity. Results indicate that our AI assistant effectively supports users in maintaining focus and aligning their digital behavior with their intentions. Our source code is publicly available at https://intentassistant.github.io2026JCJuheon Choi et al.KAISTHuman-LLM CollaborationAI-Assisted Decision-Making & AutomationSmartphone Addiction & Digital WellbeingCHI
TwistLens: A Docent-Informed Image Transformation to Create Previews That Prompt Anticipation and Interpretive Experiences Before Museum VisitsPre-visit information can enrich museum experiences, yet creates a dilemma: text-only descriptions can overwhelm without visual anchors, while viewing artworks in advance can spoil surprise. To address this tension, we introduce TwistLens, a docent-informed, AI-supported image transformation system that generates twisted previews--transformed images that convey interpretive cues while concealing original visuals. TwistLens extracts key cues from docent text using a structured taxonomy, then applies two strategies: EchoLens, which preserves intended description while altering representation, and DecoyLens, which distorts described information while maintaining representational coherence. A co-design study identified strategy preferences by information type, informing category-specific refinements. A controlled evaluation further showed that TwistLens preserves anticipation, triggers curiosity, and supports active learning without visual spoil. These findings demonstrate how semantically-aware image transformation can balance knowledge delivery and anticipation in museum contexts.2026TVThao Phuong Vu et al.Yonsei UniversityMuseum & Cultural Heritage DigitizationTangible User Interface DesignPhysical-Digital Hybrid InteractionCHI
Transformer Explainer: Learning LLM Transformers with Interactive Visual Explanation and ExperimentationThe Transformer architecture underpins modern large language models powering state-of-the-art text generation and AI applications. However, its complexity makes it difficult for non-experts to learn. Existing resources often lack interactivity, rely on static descriptions of simplified architectures, or fail to reflect models’ behavior with real data. To address this gap, we introduce Transformer Explainer, an interactive visualization tool for non-experts to learn Transformers. The tool integrates an overview illustrating the Transformer's data flow with on-demand explanations that gradually reveal mathematical details. Smooth transitions across abstraction levels highlight the interplay between high-level structures and low-level operations. Running a live GPT-2 instance directly in the browser, Transformer Explainer empowers learners to experiment with custom input and hyperparameters without setup, observing next-token predictions in real time. A 90-participant user study showed that our tool offered significant advantages in improving user understanding and engagement. Transformer Explainer has attracted over 490,000 users.2026ACAeree Cho et al.Georgia Institute of TechnologyGenerative AI (Text, Image, Music, Video)Interactive Data VisualizationPrototyping & User TestingCHI
Unpacking Visual Metaphors in Infographics: A Design SpaceVisual metaphors illuminate infographics by leveraging graphical representations from more familiar source domains (e.g., a dandelion) to explain concepts in more abstract target domains (e.g., information propagation). However, designing effective visual metaphors remains a challenge, especially for novice designers, because it requires selecting a suitable source concept for the target concept and devising a reconstruction strategy that maps the source concept to the target concept. Through a systematic review of 2,029 metaphoric infographics, we derive a design space that characterizes visual metaphors across three dimensions: target, source, and reconstruction strategy. We demonstrate the utility of our design space by transforming it into actionable design knowledge for prompting generative models in metaphor ideation. A user study with 30 participants shows that design-space-augmented prompting generates more diverse and inspiring metaphor designs than direct prompting without design-space cues.2026YGYukai Guo et al.Tsinghua UniversityData StorytellingInteractive Data VisualizationCHI
Challenges in Synchronous & Remote Collaboration Around VisualizationWe characterize 16 challenges faced by those investigating and developing remote and synchronous collaborative experiences around visualization. Our work reflects the perspectives and prior research efforts of an international group of 29 experts from across human-computer interaction and visualization sub-communities. The challenges are anchored around five collaborative activities that exhibit a centrality of visualization and multimodal communication. These activities include exploratory data analysis, creative ideation, visualization-rich presentations, joint decision making grounded in data, and real-time data monitoring. The challenges also reflect the changing dynamics of these activities in the face of recent advances in extended reality (XR) and artificial intelligence (AI). As an organizing scheme for future research at the intersection of visualization and computer-supported cooperative work, we align the challenges with a sequence of four sets of research and development activities: technological choices, social factors, AI assistance, and evaluation.2026MBMatthew Brehmer et al.University of WaterlooInteractive Data VisualizationRemote Work Tools & ExperienceMulti-User Large Display CollaborationCHI
Envisioning Posthuman/More-than-Human Futures for XR through HCI: A Critical ReorientationExtended Reality (XR) is uniquely capable of spatial and embodied simulations for various modes of worlds, beings, and relations, which makes it a medium especially well suited to exploring, realizing, and advancing posthuman/more-than-human HCI agendas. However, the three major domains of HCI research on XR---space, body, and interaction---have largely been shaped by anthropocentric design assumptions that prioritize perceptual fidelity, human-like representation, and instrumental control. This paper argues for a critical reorientation beyond these norms by shifting the direction for the domains: the design of XR environments as speculative worlding; embodiments as distributed decentering; and enactments as entangled becoming. It further recognizes XR and posthuman/more-than-human HCI to be mutually enabling, and proposes a co-constitutive framework where they can both evolve as the three domains continuously shape one another. Its contribution is a conceptual provocation that provides a diagnostic vocabulary and schema to reinvent XR as a site for cultivating speculative, inclusive, and transformative futures through HCI research.2026JSJae-eun ShinYonsei UniversityImmersion & Presence ResearchHuman-Nature Relationships (More-than-Human Design)Technology Ethics & Critical HCICHI
Toward Independent Online Shopping of the Visually Impaired Through Voice-based Computer-Using AgentVisually impaired individuals face barriers in online shopping because product details are often conveyed through images, and alternative text is frequently insufficient. The recent advent of Computer-Using Agents (CUA) based on Large Multimodal Models, which can directly manipulate graphical user interfaces, offers new opportunities for such accessibility. However, there is a lack of research that considers how voice-based systems should be designed to support visually impaired users in complex online shopping contexts. Thus, our study qualitatively explores the experiences and needs of visually impaired users as they shop online through voice interaction with a CUA We conducted a Semi-Automatic Wizard-of-Oz study with 12 visually impaired participants, in which they performed shopping tasks with a voice-based CUA system, followed by debriefing interviews. This paper proposes design implications for creating a more inclusive and disability-centered online shopping environment with voice-based CUA for visually impaired users.2026SSSubin Shin et al.Yonsei UniversityVoice AccessibilityVoice User Interface (VUI) DesignAI-Assisted Decision-Making & AutomationCHI