Searcharxiv⌕ Search

arXiv subjects

Zahra Ashktorab

Publications and source records attributed to Zahra Ashktorab.

At least 19 recordsLinked to original sources
arXiv2026

MultEval: Supporting Collaborative Alignment for LLM-as-a-Judge Evaluation Criteria

LLM-as-a-judge approaches have emerged as a scalable solution for evaluating model behaviors, yet they rely on evaluation criteria often created by a single individual, embedding that person's assumptions, priorities, and interpretive lens. In practice, defining such criteria is a collaborative and contested process involving multiple stakeholders with different values, interpretations, and priorities; an aspect largely unsupported by existing tools. To examine this problem in depth, we present a formative study examining how stakeholders collaboratively create, negotiate, and refine evaluation criteria for LLM-as-a-judge systems. Our findings reveal challenges in human oversight, including difficulties in establishing shared understanding, aligning values across stakeholders with different expertise and priorities, and translating nuanced human judgments into criteria that are interpretable and actionable for LLM judges. Based on these insights, we developed MultEval, a system that supports collaborative criteria by enabling multiple evaluators to surface and diagnose disagreements using consensus-building theory, iteratively revise criteria with attached examples and proposal history, and maintain transparency over how judgments are encoded into an automated evaluator. We further report a case study in which a team of domain experts used MultEval to collaboratively author criteria, illustrating how coordination and collaborative consensus-making shape criteria evolution.

cs.HC↗
arXiv2026

Togedule: Scheduling Meetings with Large Language Models and Adaptive Representations of Group Availability

Scheduling is a perennial-and often challenging-problem for many groups. Existing tools are mostly static, showing an identical set of choices to everyone, regardless of the current status of attendees' inputs and preferences. In this paper, we propose Togedule, an adaptive scheduling tool that uses large language models to dynamically adjust the pool of choices and their presentation format. With the initial prototype, we conducted a formative study (N=10) and identified the potential benefits and risks of such an adaptive scheduling tool. Then, after enhancing the system, we conducted two controlled experiments, one each for attendees and organizers (total N=66). For each experiment, we compared scheduling with verbal messages, shared calendars, or Togedule. Results show that Togedule significantly reduces the cognitive load of attendees indicating their availability and improves the speed and quality of the decisions made by organizers.

cs.HC↗
arXiv2026

The Behavioral Fabric of LLM-Powered GUI Agents: Human Values and Interaction Outcomes

Large Language Model (LLM)-powered web GUI agents are increasingly automating everyday online tasks. Despite their popularity, little is known about how users' preferences and values impact agents' reasoning and behavior. In this work, we investigate how both explicit and implicit user preferences, as well as the underlying user values, influence agent decision-making and action trajectories. We built a controlled testbed of 14 common interactive web tasks, spanning shopping, travel, dining, and housing, each replicated from real websites and integrated with a low-fidelity LLM-based recommender system. We injected 12 human preferences and values as personas into four state-of-the-art agents and systematically analyzed their task behaviors. Our results show that preference and value-infused prompts consistently guided agents toward outcomes that reflected these preferences and values. While the absence of user preference or value guidance led agents to exhibit a strong efficiency bias and employ shortest-path strategies, their presence steered agents' behavior trajectories through the greater use of corresponding filters and interactive web features. Despite their influence, dominant interface cues, such as discounts and advertisements, frequently overrode these effects, shortening the agents' action trajectories and inducing rationalizations that masked rather than reflected value-consistent reasoning. The contributions of this paper are twofold: (1) an open-source testbed for studying the influence of values in agent behaviors, and (2) an empirical investigation of how user preferences and values shape web agent behaviors.

cs.HC↗
arXiv2025

Generate, Evaluate, Iterate: Synthetic Data for Human-in-the-Loop Refinement of LLM Judges

The LLM-as-a-judge paradigm enables flexible, user-defined evaluation, but its effectiveness is often limited by the scarcity of diverse, representative data for refining criteria. We present a tool that integrates synthetic data generation into the LLM-as-a-judge workflow, empowering users to create tailored and challenging test cases with configurable domains, personas, lengths, and desired outcomes, including borderline cases. The tool also supports AI-assisted inline editing of existing test cases. To enhance transparency and interpretability, it reveals the prompts and explanations behind each generation. In a user study (N=24), 83% of participants preferred the tool over manually creating or selecting test cases, as it allowed them to rapidly generate diverse synthetic data without additional workload. The generated synthetic data proved as effective as hand-crafted data for both refining evaluation criteria and aligning with human preferences. These findings highlight synthetic data as a promising alternative, particularly in contexts where efficiency and scalability are critical.

cs.HC↗
arXiv2025

EvalAssist: A Human-Centered Tool for LLM-as-a-Judge

With the broad availability of large language models and their ability to generate vast outputs using varied prompts and configurations, determining the best output for a given task requires an intensive evaluation process, one where machine learning practitioners must decide how to assess the outputs and then carefully carry out the evaluation. This process is both time-consuming and costly. As practitioners work with an increasing number of models, they must now evaluate outputs to determine which model and prompt performs best for a given task. LLMs are increasingly used as evaluators to filter training data, evaluate model performance, assess harms and risks, or assist human evaluators with detailed assessments. We present EvalAssist, a framework that simplifies the LLM-as-a-judge workflow. The system provides an online criteria development environment, where users can interactively build, test, and share custom evaluation criteria in a structured and portable format. We support a set of LLM-based evaluation pipelines that leverage off-the-shelf LLMs and use a prompt-chaining approach we developed and contributed to the UNITXT open-source library. Additionally, our system also includes specially trained evaluators to detect harms and risks in LLM outputs. We have deployed the system internally in our organization with several hundreds of users.

cs.HC↗
arXiv2025

A Risk Taxonomy and Reflection Tool for Large Language Model Adoption in Public Health

Recent breakthroughs in large language models (LLMs) have generated both interest and concern about their potential adoption as information sources or communication tools across different domains. In public health, where stakes are high and impacts extend across diverse populations, adopting LLMs poses unique challenges that require thorough evaluation. However, structured approaches for assessing potential risks in public health remain under-explored. To address this gap, we conducted focus groups with public health professionals and individuals with lived experience to unpack their concerns, situated across three distinct and critical public health issues that demand high-quality information: infectious disease prevention (vaccines), chronic and well-being care (opioid use disorder), and community health and safety (intimate partner violence). We synthesize participants' perspectives into a risk taxonomy, identifying and contextualizing the potential harms LLMs may introduce when positioned alongside traditional health communication. This taxonomy highlights four dimensions of risk to individuals, human-centered care, information ecosystem, and technology accountability. For each dimension, we unpack specific risks and offer example reflection questions to help practitioners adopt a risk-reflexive approach. By summarizing distinctive LLM characteristics and linking them to identified risks, we discuss the need to revisit prior mental models of information behaviors and complement evaluations with external validity and domain expertise through lived experience and real-world practices. Together, this work contributes a shared vocabulary and reflection tool for people in both computing and public health to collaboratively anticipate, evaluate, and mitigate risks in deciding when to employ LLM capabilities (or not) and how to mitigate harm.

cs.HC↗
arXiv2025

Aligning Human and LLM Judgments: Insights from EvalAssist on Task-Specific Evaluations and AI-assisted Assessment Strategy Preferences

Evaluation of large language model (LLM) outputs requires users to make critical judgments about the best outputs across various configurations. This process is costly and takes time given the large amounts of data. LLMs are increasingly used as evaluators to filter training data, evaluate model performance or assist human evaluators with detailed assessments. To support this process, effective front-end tools are critical for evaluation. Two common approaches for using LLMs as evaluators are direct assessment and pairwise comparison. In our study with machine learning practitioners (n=15), each completing 6 tasks yielding 131 evaluations, we explore how task-related factors and assessment strategies influence criteria refinement and user perceptions. Findings show that users performed more evaluations with direct assessment by making criteria task-specific, modifying judgments, and changing the evaluator model. We conclude with recommendations for how systems can better support interactions in LLM-assisted evaluations.

cs.HC↗
arXiv2025

Who's Sorry Now: User Preferences Among Rote, Empathic, and Explanatory Apologies from LLM Chatbots

As chatbots driven by large language models (LLMs) are increasingly deployed in everyday contexts, their ability to recover from errors through effective apologies is critical to maintaining user trust and satisfaction. In a preregistered study with Prolific workers (N=162), we examine user preferences for three types of apologies (rote, explanatory, and empathic) issued in response to three categories of common LLM mistakes (bias, unfounded fabrication, and factual errors). We designed a pairwise experiment in which participants evaluated chatbot responses consisting of an initial error, a subsequent apology, and a resolution. Explanatory apologies were generally preferred, but this varied by context and user. In the bias scenario, empathic apologies were favored for acknowledging emotional impact, while hallucinations, though seen as serious, elicited no clear preference, reflecting user uncertainty. Our findings show the complexity of effective apology in AI systems. We discuss key insights such as personalization and calibration that future systems must navigate to meaningfully repair trust.

cs.HC↗
arXiv2025

Position: Theory of Mind Benchmarks are Broken for Large Language Models

Our paper argues that the majority of theory of mind benchmarks are broken because of their inability to directly test how large language models (LLMs) adapt to new partners. This problem stems from the fact that theory of mind benchmarks for LLMs are overwhelmingly inspired by the methods used to test theory of mind in humans and fall victim to a fallacy of attributing human-like qualities to AI agents. We expect that humans will engage in a consistent reasoning process across various questions about a situation, but this is known to not be the case for current LLMs. Most theory of mind benchmarks only measure what we call literal theory of mind: the ability to predict the behavior of others. However, this type of metric is only informative when agents exhibit self-consistent reasoning. Thus, we introduce the concept of functional theory of mind: the ability to adapt to agents in-context following a rational response to their behavior. We find that many open source LLMs are capable of displaying strong literal theory of mind capabilities, but seem to struggle with functional theory of mind -- even with exceedingly simple partner policies. Simply put, strong literal theory of mind performance does not necessarily imply strong functional theory of mind performance or vice versa. Achieving functional theory of mind, particularly over long interaction horizons with a partner, is a significant challenge deserving a prominent role in any meaningful LLM theory of mind evaluation.

cs.AI↗
arXiv2025

A Case Study Investigating the Role of Generative AI in Quality Evaluations of Epics in Agile Software Development

The broad availability of generative AI offers new opportunities to support various work domains, including agile software development. Agile epics are a key artifact for product managers to communicate requirements to stakeholders. However, in practice, they are often poorly defined, leading to churn, delivery delays, and cost overruns. In this industry case study, we investigate opportunities for large language models (LLMs) to evaluate agile epic quality in a global company. Results from a user study with 17 product managers indicate how LLM evaluations could be integrated into their work practices, including perceived values and usage in improving their epics. High levels of satisfaction indicate that agile epics are a new, viable application of AI evaluations. However, our findings also outline challenges, limitations, and adoption barriers that can inform both practitioners and researchers on the integration of such evaluations into future agile work practices.

cs.SE↗
arXiv2025

Emerging Reliance Behaviors in Human-AI Content Grounded Data Generation: The Role of Cognitive Forcing Functions and Hallucinations

We investigate the impact of hallucinations and Cognitive Forcing Functions in human-AI collaborative content-grounded data generation, focusing on the use of Large Language Models (LLMs) to assist in generating high quality conversational data. Through a study with 34 users who each completed 8 tasks (n=272), we found that hallucinations significantly reduce data quality. While Cognitive Forcing Functions do not always alleviate these effects, their presence influences how users integrate AI responses. Specifically, we observed emerging reliance behaviors, with users often appending AI-generated responses to their correct answers, even when the AI's suggestions conflicted. This points to a potential drawback of Cognitive Forcing Functions, particularly when AI suggestions are inaccurate. Users who overrelied on AI-generated text produced lower quality data, emphasizing the nuanced dynamics of overreliance in human-LLM collaboration compared to traditional human-AI decision-making.

cs.HC↗
arXiv2025

Helping the Helper: Supporting Peer Counselors via AI-Empowered Practice and Feedback

Millions of users come to online peer counseling platforms to seek support. However, studies show that online peer support groups are not always as effective as expected, largely due to users' negative experiences with unhelpful counselors. Peer counselors are key to the success of online peer counseling platforms, but most often do not receive appropriate training.Hence, we introduce CARE: an AI-based tool to empower and train peer counselors through practice and feedback. Concretely, CARE helps diagnose which counseling strategies are needed in a given situation and suggests example responses to counselors during their practice sessions. Building upon the Motivational Interviewing framework, CARE utilizes large-scale counseling conversation data with text generation techniques to enable these functionalities. We demonstrate the efficacy of CARE by performing quantitative evaluations and qualitative user studies through simulated chats and semi-structured interviews, finding that CARE especially helps novice counselors in challenging situations. The code is available at https://github.com/SALT-NLP/CARE

cs.HC↗
arXiv2025

Scopes of Alignment

Much of the research focus on AI alignment seeks to align large language models and other foundation models to the context-less and generic values of helpfulness, harmlessness, and honesty. Frontier model providers also strive to align their models with these values. In this paper, we motivate why we need to move beyond such a limited conception and propose three dimensions for doing so. The first scope of alignment is competence: knowledge, skills, or behaviors the model must possess to be useful for its intended purpose. The second scope of alignment is transience: either semantic or episodic depending on the context of use. The third scope of alignment is audience: either mass, public, small-group, or dyadic. At the end of the paper, we use the proposed framework to position some technologies and workflows that go beyond prevailing notions of alignment.

cs.CY↗
arXiv2024

Granite Guardian

We introduce the Granite Guardian models, a suite of safeguards designed to provide risk detection for prompts and responses, enabling safe and responsible use in combination with any large language model (LLM). These models offer comprehensive coverage across multiple risk dimensions, including social bias, profanity, violence, sexual content, unethical behavior, jailbreaking, and hallucination-related risks such as context relevance, groundedness, and answer relevance for retrieval-augmented generation (RAG). Trained on a unique dataset combining human annotations from diverse sources and synthetic data, Granite Guardian models address risks typically overlooked by traditional risk detection models, such as jailbreaks and RAG-specific issues. With AUC scores of 0.871 and 0.854 on harmful content and RAG-hallucination-related benchmarks respectively, Granite Guardian is the most generalizable and competitive model available in the space. Released as open-source, Granite Guardian aims to promote responsible AI development across the community. https://github.com/ibm-granite/granite-guardian

cs.CL↗
arXiv2024

Black-box Uncertainty Quantification Method for LLM-as-a-Judge

LLM-as-a-Judge is a widely used method for evaluating the performance of Large Language Models (LLMs) across various tasks. We address the challenge of quantifying the uncertainty of LLM-as-a-Judge evaluations. While uncertainty quantification has been well-studied in other domains, applying it effectively to LLMs poses unique challenges due to their complex decision-making capabilities and computational demands. In this paper, we introduce a novel method for quantifying uncertainty designed to enhance the trustworthiness of LLM-as-a-Judge evaluations. The method quantifies uncertainty by analyzing the relationships between generated assessments and possible ratings. By cross-evaluating these relationships and constructing a confusion matrix based on token probabilities, the method derives labels of high or low uncertainty. We evaluate our method across multiple benchmarks, demonstrating a strong correlation between the accuracy of LLM evaluations and the derived uncertainty scores. Our findings suggest that this method can significantly improve the reliability and consistency of LLM-as-a-Judge evaluations.

cs.LG↗
arXiv2024

Human-Centered Design Recommendations for LLM-as-a-Judge

Traditional reference-based metrics, such as BLEU and ROUGE, are less effective for assessing outputs from Large Language Models (LLMs) that produce highly creative or superior-quality text, or in situations where reference outputs are unavailable. While human evaluation remains an option, it is costly and difficult to scale. Recent work using LLMs as evaluators (LLM-as-a-judge) is promising, but trust and reliability remain a significant concern. Integrating human input is crucial to ensure criteria used to evaluate are aligned with the human's intent, and evaluations are robust and consistent. This paper presents a user study of a design exploration called EvaluLLM, that enables users to leverage LLMs as customizable judges, promoting human involvement to balance trust and cost-saving potential with caution. Through interviews with eight domain experts, we identified the need for assistance in developing effective evaluation criteria aligning the LLM-as-a-judge with practitioners' preferences and expectations. We offer findings and design recommendations to optimize human-assisted LLM-as-judge systems.

cs.HC↗
arXiv2023

Fairness Evaluation in Text Classification: Machine Learning Practitioner Perspectives of Individual and Group Fairness

Mitigating algorithmic bias is a critical task in the development and deployment of machine learning models. While several toolkits exist to aid machine learning practitioners in addressing fairness issues, little is known about the strategies practitioners employ to evaluate model fairness and what factors influence their assessment, particularly in the context of text classification. Two common approaches of evaluating the fairness of a model are group fairness and individual fairness. We run a study with Machine Learning practitioners (n=24) to understand the strategies used to evaluate models. Metrics presented to practitioners (group vs. individual fairness) impact which models they consider fair. Participants focused on risks associated with underpredicting/overpredicting and model sensitivity relative to identity token manipulations. We discover fairness assessment strategies involving personal experiences or how users form groups of identity tokens to test model fairness. We provide recommendations for interactive tools for evaluating fairness in text classification.

cs.HC↗