arXiv subjects

Vikram Mohanty

Publications and source records attributed to Vikram Mohanty.

14 recordsLinked to original sources

arXiv2026

StructSim: Measuring Idea Similarity at Scale Through Structural Representation

Measuring idea similarity is fundamental to creativity evaluation, especially as LLMs enable idea generation at increasing scale. However, text embeddings collapse an idea into a single vector, making it difficult to capture structural similarity, including partial overlap across core and supporting components and differences across levels of abstraction. We introduce a shared structural representation that decomposes ideas into purpose, mechanism, and implementation components and organizes related components in a multi-layer concept graph. From this representation, we define measures of pairwise similarity and set-level mechanism coverage for assessing idea diversity. We evaluate our approach using controlled idea triples and assessments from 12 experts, focusing on differences in core mechanisms, implementations, and supporting components. Our method improves alignment with expert judgments of structural similarity by 31% over the embedding baseline and better reflects expert assessments of idea set coverage, supporting scalable evaluation of idea similarity and diversity.

cs.HC↗

arXiv2026

SleuthTalk: Supporting Historical Photo Identification with Private Workspaces for Collective Sensemaking and Deliberation

Identifying individuals in historical photographs is a critical task across fields such as history, journalism, genealogy, and archival research. While AI-based facial recognition can efficiently generate candidate matches, it often produces ambiguous results that require deeper analysis and contextual interpretation. Existing platforms lack robust support for collaborative deliberation, especially in uncertain or high-stakes cases. We present SleuthTalk, a private collaborative workspace integrated into Civil War Photo Sleuth, designed to scaffold structured comparison, discussion, and group decision-making. SleuthTalk enables users to curate custom shortlists, annotate facial features, and build consensus through structured feedback. In a mixed-methods evaluation with experienced historical photo researchers, SleuthTalk enhanced self-reported confidence, surfaced diverse perspectives, and supported transparent, reflective identifications.

cs.HC↗

arXiv2026

What to Make Sense of in the Era of LLM? A Perspective from the Structure and Efforts in Sensemaking

Sensemaking tasks often entail navigating through complex, ambiguous data to construct coherent insights. Prior work has shown that crowds can effectively distribute cognitive load, pooling diverse perspectives to enhance analytical depth. Recent advancements in LLMs have further expanded the toolkit for sensemaking, offering scalable data processing, complex pattern recognition, and the ability to infer and propose meaningful hypotheses. In this study, we explore how LLMs (i.e., GPT-4) can assist in a complex sensemaking task of deciphering fictional terrorist plots. We explore two different approaches for leveraging GPT-4's capabilities: a holistic sensemaking process and a step-by-step approach. Our preliminary investigations open the doors for future research into optimizing human-AI collaborative workflows, aiming to harness the complementary strengths of both for more effective sensemaking in complex scenarios.

cs.HC↗

arXiv2025

FlexMind: Supporting Deeper Creative Thinking with LLMs

Effective ideation requires both broad exploration of diverse ideas and deep evaluation of their potential. Generative AI can support such processes, but current tools typically emphasize either generating many ideas or supporting in-depth consideration of a few, lacking support for both. Research also highlights risks of over-reliance on LLMs, including shallow exploration and negative creative outcomes. We present FlexMind, an AI-augmented system that scaffolds iterative exploration of ideas, tradeoffs, and mitigations. FlexMind exposes users to a broad set of ideas while enabling a lightweight transition into deeper engagement. In a study comparing ideation with FlexMind to ChatGPT, participants generated higher-quality ideas with FlexMind, due to both broader exposure and deeper engagement with tradeoffs. By scaffolding ideation across breadth, depth, and reflective evaluation, FlexMind empowers users to surface ideas that might otherwise go unnoticed or be prematurely discarded.

cs.HC↗

arXiv2025

Scaffolding Flexible Ideation Workflows with AI in Creative Problem-Solving

Divergent thinking in the ideation stage of creative problem-solving demands that individuals explore a broad design space. Yet this exploration rarely follows a neat, linear sequence; problem-solvers constantly shift among searching, creating, and evaluating ideas. Existing interfaces either impose rigid, step-by-step workflows or permit unguided free-form exploration. To strike a balance between flexibility and guidance for augmenting people's efficiency and creativity, we introduce a human-AI collaborative workflow that supports a fluid ideation process. The system surfaces three opt-in aids: (1) high-level schemas to uncover alternative ideas, (2) risk analysis with mitigation suggestions, and (3) steering system-generated suggestions. Users can invoke these supports at any moment, allowing seamless back-and-forth movement among design actions to maintain creative momentum.

cs.HC↗

arXiv2025

VISLIX: An XAI Framework for Validating Vision Models with Slice Discovery and Analysis

Real-world machine learning models require rigorous evaluation before deployment, especially in safety-critical domains like autonomous driving and surveillance. The evaluation of machine learning models often focuses on data slices, which are subsets of the data that share a set of characteristics. Data slice finding automatically identifies conditions or data subgroups where models underperform, aiding developers in mitigating performance issues. Despite its popularity and effectiveness, data slicing for vision model validation faces several challenges. First, data slicing often needs additional image metadata or visual concepts, and falls short in certain computer vision tasks, such as object detection. Second, understanding data slices is a labor-intensive and mentally demanding process that heavily relies on the expert's domain knowledge. Third, data slicing lacks a human-in-the-loop solution that allows experts to form hypothesis and test them interactively. To overcome these limitations and better support the machine learning operations lifecycle, we introduce VISLIX, a novel visual analytics framework that employs state-of-the-art foundation models to help domain experts analyze slices in computer vision models. Our approach does not require image metadata or visual concepts, automatically generates natural language insights, and allows users to test data slice hypothesis interactively. We evaluate VISLIX with an expert study and three use cases, that demonstrate the effectiveness of our tool in providing comprehensive insights for validating object detection models.

cs.CV↗

arXiv2025

From Overload to Insight: Scaffolding Creative Ideation through Structuring Inspiration

Creative ideation relies on exploring diverse stimuli, but the overwhelming abundance of information often makes it difficult to identify valuable insights or reach the `aha' moment. Traditional methods for accessing design stimuli lack organization and fail to support users in discovering promising opportunities within large idea spaces. In this position paper, we explore how AI can be leveraged to structure, organize, and surface relevant stimuli, guiding users in both exploring idea spaces and mapping insights back to their design challenges.

cs.HC↗

arXiv2025

InterChat: Enhancing Generative Visual Analytics using Multimodal Interactions

The rise of Large Language Models (LLMs) and generative visual analytics systems has transformed data-driven insights, yet significant challenges persist in accurately interpreting users' analytical and interaction intents. While language inputs offer flexibility, they often lack precision, making the expression of complex intents inefficient, error-prone, and time-intensive. To address these limitations, we investigate the design space of multimodal interactions for generative visual analytics through a literature review and pilot brainstorming sessions. Building on these insights, we introduce a highly extensible workflow that integrates multiple LLM agents for intent inference and visualization generation. We develop InterChat, a generative visual analytics system that combines direct manipulation of visual elements with natural language inputs. This integration enables precise intent communication and supports progressive, visually driven exploratory data analyses. By employing effective prompt engineering, and contextual interaction linking, alongside intuitive visualization and interaction designs, InterChat bridges the gap between user interactions and LLM-driven visualizations, enhancing both interpretability and usability. Extensive evaluations, including two usage scenarios, a user study, and expert feedback, demonstrate the effectiveness of InterChat. Results show significant improvements in the accuracy and efficiency of handling complex visual analytics tasks, highlighting the potential of multimodal interactions to redefine user engagement and analytical depth in generative visual analytics.

cs.HC↗

arXiv2025

Demystifying CO2: lessons from nutrition labeling and step counting

There is growing concern about climate change and increased interest in taking action. However, people have difficulty understanding abstract units like CO2 and the relative environmental impact of different behaviors. This position piece explores findings from nutritional labeling and step counting research, two domains aimed at making abstract concepts (i.e., calories and exercise) more familiar to the general public. Research in these two domains suggests that consistent, widespread communication can make people more familiar and think more precisely about abstract units, but that better communication and understanding does not guarantee behavior change. These findings suggest that consistent and ubiquitous communication can make CO2 units more familiar to people, which in turn could help interventions aimed at encouraging more sustainable behaviors.

cs.HC↗

arXiv2025

What Lies Beneath? Exploring the Impact of Underlying AI Model Updates in AI-Infused Systems

AI models are constantly evolving, with new versions released frequently. Human-AI interaction guidelines encourage notifying users about changes in model capabilities, ideally supported by thorough benchmarking. However, as AI systems integrate into domain-specific workflows, exhaustive benchmarking can become impractical, often resulting in silent or minimally communicated updates. This raises critical questions: Can users notice these updates? What cues do they rely on to distinguish between models? How do such changes affect their behavior and task performance? We address these questions through two studies in the context of facial recognition for historical photo identification: an online experiment examining users' ability to detect model updates, followed by a diary study exploring perceptions in a real-world deployment. Our findings highlight challenges in noticing AI model updates, their impact on downstream user behavior and performance, and how they lead users to develop divergent folk theories. Drawing on these insights, we discuss strategies for effectively communicating model updates in AI-infused systems.

cs.HC↗

arXiv2024

Save A Tree or 6 kg of CO2? Understanding Effective Carbon Footprint Interventions for Eco-Friendly Vehicular Choices

From ride-hailing to car rentals, consumers are often presented with eco-friendly options. Beyond highlighting a "green" vehicle and CO2 emissions, CO2 equivalencies have been designed to provide understandable amounts; we ask which equivalencies will lead to eco-friendly decisions. We conducted five ride-hailing scenario surveys where participants picked between regular and eco-friendly options, testing equivalencies, social features, and valence-based interventions. Further, we tested a car-rental embodiment to gauge how an individual (needing a car for several days) might behave versus the immediate ride-hailing context. We find that participants are more likely to choose green rides when presented with additional information about emissions; CO2 by weight was found to be the most effective. Further, we found that information framing - be it individual or collective footprint, positive or negative valence - had an impact on participants' choices. Finally, we discuss how our findings inform the design of effective interventions for reducing car-based carbon-emissions.

cs.HC↗

arXiv2024

Investigating Interaction Modes and User Agency in Human-LLM Collaboration for Domain-Specific Data Analysis

Despite demonstrating robust capabilities in performing tasks related to general-domain data-operation tasks, Large Language Models (LLMs) may exhibit shortcomings when applied to domain-specific tasks. We consider the design of domain-specific AI-powered data analysis tools from two dimensions: interaction and user agency. We implemented two design probes that fall on the two ends of the two dimensions: an open-ended high agency (OHA) prototype and a structured low agency (SLA) prototype. We conducted an interview study with nine data scientists to investigate (1) how users perceived the LLM outputs for data analysis assistance, and (2) how the two test design probes, OHA and SLA, affected user behavior, performance, and perceptions. Our study revealed insights regarding participants' interactions with LLMs, how they perceived the results, and their desire for explainability concerning LLM outputs, along with a noted need for collaboration with other users, and how they envisioned the utility of LLMs in their workflow.

cs.HC↗

arXiv2023

DoubleCheck: Designing Community-based Assessability for Historical Person Identification

Historical photos are valuable for their cultural and economic significance, but can be difficult to identify accurately due to various challenges such as low-quality images, lack of corroborating evidence, and limited research resources. Misidentified photos can have significant negative consequences, including lost economic value, incorrect historical records, and the spread of misinformation that can lead to perpetuating conspiracy theories. To accurately assess the credibility of a photo identification (ID), it may be necessary to conduct investigative research, use domain knowledge, and consult experts. In this paper, we introduce DoubleCheck, a quality assessment framework for verifying historical photo IDs on Civil War Photo Sleuth (CWPS), a popular online platform for identifying American Civil War-era photos using facial recognition and crowdsourcing. DoubleCheck focuses on improving CWPS's user experience and system architecture to display information useful for assessing the quality of historical photo IDs on CWPS. In a mixed-methods evaluation of DoubleCheck, we found that users contributed a wide diversity of sources for photo IDs, which helped facilitate the community's assessment of these IDs through DoubleCheck's provenance visualizations. Further, DoubleCheck's quality assessment badges and visualizations supported users in making accurate assessments of photo IDs, even in cases involving ID conflicts.

cs.HC↗

arXiv2016

DeepVO: A Deep Learning approach for Monocular Visual Odometry

Deep Learning based techniques have been adopted with precision to solve a lot of standard computer vision problems, some of which are image classification, object detection and segmentation. Despite the widespread success of these approaches, they have not yet been exploited largely for solving the standard perception related problems encountered in autonomous navigation such as Visual Odometry (VO), Structure from Motion (SfM) and Simultaneous Localization and Mapping (SLAM). This paper analyzes the problem of Monocular Visual Odometry using a Deep Learning-based framework, instead of the regular 'feature detection and tracking' pipeline approaches. Several experiments were performed to understand the influence of a known/unknown environment, a conventional trackable feature and pre-trained activations tuned for object classification on the network's ability to accurately estimate the motion trajectory of the camera (or the vehicle). Based on these observations, we propose a Convolutional Neural Network architecture, best suited for estimating the object's pose under known environment conditions, and displays promising results when it comes to inferring the actual scale using just a single camera in real-time.

cs.CV↗