Crystal Qian
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Crystal Qian has a total of 31 co-authors. Here are the top co-authors by number of shared publications:
- Emily Reif
Google research
5 shared publications - James Wexler
Google Research, United States
5 shared publications - James Wexler
Harvard University MIT & NBER MIT
3 shared publications - Minsuk Kahng
Yonsei University, Seoul, Republic of Korea
3 shared publications - Minsuk Kahng
Google research
2 shared publications - Nithum Thain
Google Deepmind, Mountain View, USA
2 shared publications
Latest Publications
- Open Access
- 25 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 25-59
Real-Time Group Dynamics with LLM Facilitation: Evidence from a Charity Allocation Task
- 25 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 25-59
Open AccessReal-Time Group Dynamics with LLM Facilitation: Evidence from a Charity Allocation Task
- Open Access
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1625-1646
Strategic Tradeoffs Between Humans and AI in Multi-Agent Bargaining
- 22 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1625-1646
Open AccessStrategic Tradeoffs Between Humans and AI in Multi-Agent Bargaining
- Open Access
- 15 January 2026
- Published by Center for Open Science in PsyArXiv
A Moral Turing Test to assess How Subjective Belief and Objective Source Affect Detection and Agreement with LLM Judgments
As large language models (LLMs) are increasingly integrated into decision-making systems—such as autonomous vehicles and medical devices—understanding how humans perceive and evaluate AI-generated judgements is crucial. To investigate this, we conducted a series of experiments in which participants evaluated justifications for moral and non-moral choices, generated either by humans or LLMs. Participants attempted to identify the source of each justification (either human or LLM) and indicated their agreement with its content. We found that while detection accuracy was consistently above chance, it remained below 70%. In terms of agreement, there was no overall preference for human-generated responses, even though machine-generated justifications were favored in particularly challenging moral scenarios. Notably, we observed a systematic anti-AI bias: participants were less likely to agree with judgments they believed were AI-generated, regardless of the true source. Linguistic cues, such as response length, typos, first-person pronouns, and cost-benefit language markers (e.g., “lives,” “save”), influenced both detection and agreement. Participants tended to disagree with cost-benefit calculations, possibly due to an expectation that AI would favor such reasoning. These findings highlight the influence of motivated belief and ingroup/outgroup bias in shaping human evaluation of AI-generated content, particularly in morally sensitive contexts
- 15 January 2026
- Published by Center for Open Science in PsyArXiv
Open AccessA Moral Turing Test to assess How Subjective Belief and Objective Source Affect Detection and Agreement with LLM Judgments
As large language models (LLMs) are increasingly integrated into decision-making systems—such as autonomous vehicles and medical devices—understanding how humans perceive and evaluate AI-generated judgements is crucial. To investigate this, we conducted a series of experiments in which participants evaluated justifications for moral and non-moral choices, generated either by humans or LLMs. Participants attempted to identify the source of each justification (either human or LLM) and indicated their agreement with its content. We found that while detection accuracy was consistently above chance, it remained below 70%. In terms of agreement, there was no overall preference for human-generated responses, even though machine-generated justifications were favored in particularly challenging moral scenarios. Notably, we observed a systematic anti-AI bias: participants were less likely to agree with judgments they believed were AI-generated, regardless of the true source. Linguistic cues, such as response length, typos, first-person pronouns, and cost-benefit language markers (e.g., “lives,” “save”), influenced both detection and agreement. Participants tended to disagree with cost-benefit calculations, possibly due to an expectation that AI would favor such reasoning. These findings highlight the influence of motivated belief and ingroup/outgroup bias in shaping human evaluation of AI-generated content, particularly in morally sensitive contexts
- Open Access
Strategic Tradeoffs Between Humans and AI in Multi-Agent Bargaining
Coordination tasks traditionally performed by humans are increasingly being delegated to autonomous agents. As this pattern progresses, it becomes critical to evaluate not only these agents' performance but also the processes through which they negotiate in dynamic, multi-agent environments. Furthermore, different agents exhibit distinct advantages: traditional statistical agents, such as Bayesian models, may excel under well-specified conditions, whereas large language models (LLMs) can generalize across contexts. In this work, we compare humans (N = 216), LLMs (GPT-4o, Gemini 1.5 Pro), and Bayesian agents in a dynamic negotiation setting that enables direct, identical-condition comparisons across populations, capturing both outcomes and behavioral dynamics. Bayesian agents extract the highest surplus through aggressive optimization, at the cost of frequent trade rejections. Humans and LLMs can achieve similar overall surplus, but through distinct behaviors: LLMs favor conservative, concessionary trades with few rejections, while humans employ more strategic, risk-taking, and fairness-oriented behaviors. Thus, we find that performance parity -- a common benchmark in agent evaluation -- can conceal fundamental differences in process and alignment, which are critical for practical deployment in real-world coordination tasks.
Open AccessStrategic Tradeoffs Between Humans and AI in Multi-Agent Bargaining
Coordination tasks traditionally performed by humans are increasingly being delegated to autonomous agents. As this pattern progresses, it becomes critical to evaluate not only these agents' performance but also the processes through which they negotiate in dynamic, multi-agent environments. Furthermore, different agents exhibit distinct advantages: traditional statistical agents, such as Bayesian models, may excel under well-specified conditions, whereas large language models (LLMs) can generalize across contexts. In this work, we compare humans (N = 216), LLMs (GPT-4o, Gemini 1.5 Pro), and Bayesian agents in a dynamic negotiation setting that enables direct, identical-condition comparisons across populations, capturing both outcomes and behavioral dynamics. Bayesian agents extract the highest surplus through aggressive optimization, at the cost of frequent trade rejections. Humans and LLMs can achieve similar overall surplus, but through distinct behaviors: LLMs favor conservative, concessionary trades with few rejections, while humans employ more strategic, risk-taking, and fairness-oriented behaviors. Thus, we find that performance parity -- a common benchmark in agent evaluation -- can conceal fundamental differences in process and alignment, which are critical for practical deployment in real-world coordination tasks.
- Open Access
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-10
LLM Adoption in Data Curation Workflows: Industry Practices and Insights
- 25 April 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-10
Open AccessLLM Adoption in Data Curation Workflows: Industry Practices and Insights
- Open Access
The Moral Turing Test: Evaluating Human-LLM Alignment in Moral Decision-Making
As large language models (LLMs) become increasingly integrated into society, their alignment with human morals is crucial. To better understand this alignment, we created a large corpus of human- and LLM-generated responses to various moral scenarios. We found a misalignment between human and LLM moral assessments; although both LLMs and humans tended to reject morally complex utilitarian dilemmas, LLMs were more sensitive to personal framing. We then conducted a quantitative user study involving 230 participants (N=230), who evaluated these responses by determining whether they were AI-generated and assessed their agreement with the responses. Human evaluators preferred LLMs' assessments in moral scenarios, though a systematic anti-AI bias was observed: participants were less likely to agree with judgments they believed to be machine-generated. Statistical and NLP-based analyses revealed subtle linguistic differences in responses, influencing detection and agreement. Overall, our findings highlight the complexities of human-AI perception in morally charged decision-making.
Open AccessThe Moral Turing Test: Evaluating Human-LLM Alignment in Moral Decision-Making
As large language models (LLMs) become increasingly integrated into society, their alignment with human morals is crucial. To better understand this alignment, we created a large corpus of human- and LLM-generated responses to various moral scenarios. We found a misalignment between human and LLM moral assessments; although both LLMs and humans tended to reject morally complex utilitarian dilemmas, LLMs were more sensitive to personal framing. We then conducted a quantitative user study involving 230 participants (N=230), who evaluated these responses by determining whether they were AI-generated and assessed their agreement with the responses. Human evaluators preferred LLMs' assessments in moral scenarios, though a systematic anti-AI bias was observed: participants were less likely to agree with judgments they believed to be machine-generated. Statistical and NLP-based analyses revealed subtle linguistic differences in responses, influencing detection and agreement. Overall, our findings highlight the complexities of human-AI perception in morally charged decision-making.
- Open Access
- 11 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-9
Automatic Histograms: Leveraging Language Models for Text Dataset Exploration
Making sense of unstructured text datasets is perennially difficult, yet increasingly relevant with Large Language Models. Data practitioners often rely on dataset summaries, especially distributions of various derived features. Some features, like toxicity or topics, are relevant to many datasets, but many interesting features are domain specific: instruments and genres for a music dataset, or diseases and symptoms for a medical dataset. Accordingly, data practitioners often run custom analyses for each dataset, which is cumbersome and difficult, or use unsupervised methods. We present AutoHistograms, a visualization tool leveraging LLMs. AutoHistograms automatically identifies relevant entity-based features, visualizes them, and allows the user to interactively query the dataset for new categories of entities. In a user study with (n=10) data practitioners, we observe that participants were able to quickly onboard to AutoHistograms, use the tool to identify actionable insights, and conceptualize a broad range of applicable use cases. Together, this tool and user study contribute to the growing field of LLM-assisted sensemaking tools.
- 11 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-9
Open AccessAutomatic Histograms: Leveraging Language Models for Text Dataset Exploration
Making sense of unstructured text datasets is perennially difficult, yet increasingly relevant with Large Language Models. Data practitioners often rely on dataset summaries, especially distributions of various derived features. Some features, like toxicity or topics, are relevant to many datasets, but many interesting features are domain specific: instruments and genres for a music dataset, or diseases and symptoms for a medical dataset. Accordingly, data practitioners often run custom analyses for each dataset, which is cumbersome and difficult, or use unsupervised methods. We present AutoHistograms, a visualization tool leveraging LLMs. AutoHistograms automatically identifies relevant entity-based features, visualizes them, and allows the user to interactively query the dataset for new categories of entities. In a user study with (n=10) data practitioners, we observe that participants were able to quickly onboard to AutoHistograms, use the tool to identify actionable insights, and conceptualize a broad range of applicable use cases. Together, this tool and user study contribute to the growing field of LLM-assisted sensemaking tools.
- Open Access
- 2 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Understanding the Dataset Practitioners Behind Large Language Models
As large language models (LLMs) become more advanced and impactful, it is increasingly important to scrutinize the data that they rely upon and produce. What is it to be a dataset practitioner doing this work? We approach this in two parts: first, we define the role of “dataset practitioners” by performing a retrospective analysis on the responsibilities of teams contributing to LLM development at a technology company, Google. Then, we conduct semi-structured interviews with a cross-section of these practitioners (N=10). We find that although data quality is a top priority, there is little consensus around what data quality is and how to evaluate it. Consequently, practitioners either rely on their own intuition or write custom code to evaluate their data. We discuss potential reasons for this phenomenon and opportunities for alignment.
- 2 May 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessUnderstanding the Dataset Practitioners Behind Large Language Models
As large language models (LLMs) become more advanced and impactful, it is increasingly important to scrutinize the data that they rely upon and produce. What is it to be a dataset practitioner doing this work? We approach this in two parts: first, we define the role of “dataset practitioners” by performing a retrospective analysis on the responsibilities of teams contributing to LLM development at a technology company, Google. Then, we conduct semi-structured interviews with a cross-section of these practitioners (N=10). We find that although data quality is a top priority, there is little consensus around what data quality is and how to evaluate it. Consequently, practitioners either rely on their own intuition or write custom code to evaluate their data. We discuss potential reasons for this phenomenon and opportunities for alignment.
- Open Access
Take It, Leave It, or Fix It: Measuring Productivity and Trust in Human-AI Collaboration
Although recent developments in generative AI have greatly enhanced the capabilities of conversational agents such as Google's Gemini (formerly Bard) or OpenAI's ChatGPT, it's unclear whether the usage of these agents aids users across various contexts. To better understand how access to conversational AI affects productivity and trust, we conducted a mixed-methods, task-based user study, observing 76 software engineers (N=76) as they completed a programming exam with and without access to Bard. Effects on performance, efficiency, satisfaction, and trust vary depending on user expertise, question type (open-ended "solve" vs. definitive "search" questions), and measurement type (demonstrated vs. self-reported). Our findings include evidence of automation complacency, increased reliance on the AI over the course of the task, and increased performance for novices on "solve"-type questions when using the AI. We discuss common behaviors, design recommendations, and impact considerations to improve collaborations with conversational AI.
Open AccessTake It, Leave It, or Fix It: Measuring Productivity and Trust in Human-AI Collaboration
Although recent developments in generative AI have greatly enhanced the capabilities of conversational agents such as Google's Gemini (formerly Bard) or OpenAI's ChatGPT, it's unclear whether the usage of these agents aids users across various contexts. To better understand how access to conversational AI affects productivity and trust, we conducted a mixed-methods, task-based user study, observing 76 software engineers (N=76) as they completed a programming exam with and without access to Bard. Effects on performance, efficiency, satisfaction, and trust vary depending on user expertise, question type (open-ended "solve" vs. definitive "search" questions), and measurement type (demonstrated vs. self-reported). Our findings include evidence of automation complacency, increased reliance on the AI over the course of the task, and increased performance for novices on "solve"-type questions when using the AI. We discuss common behaviors, design recommendations, and impact considerations to improve collaborations with conversational AI.
- Open Access
Understanding the Dataset Practitioners Behind Large Language Model Development
As large language models (LLMs) become more advanced and impactful, it is increasingly important to scrutinize the data that they rely upon and produce. What is it to be a dataset practitioner doing this work? We approach this in two parts: first, we define the role of "dataset practitioners" by performing a retrospective analysis on the responsibilities of teams contributing to LLM development at a technology company, Google. Then, we conduct semi-structured interviews with a cross-section of these practitioners (N=10). We find that although data quality is a top priority, there is little consensus around what data quality is and how to evaluate it. Consequently, practitioners either rely on their own intuition or write custom code to evaluate their data. We discuss potential reasons for this phenomenon and opportunities for alignment.
Open AccessUnderstanding the Dataset Practitioners Behind Large Language Model Development
As large language models (LLMs) become more advanced and impactful, it is increasingly important to scrutinize the data that they rely upon and produce. What is it to be a dataset practitioner doing this work? We approach this in two parts: first, we define the role of "dataset practitioners" by performing a retrospective analysis on the responsibilities of teams contributing to LLM development at a technology company, Google. Then, we conduct semi-structured interviews with a cross-section of these practitioners (N=10). We find that although data quality is a top priority, there is little consensus around what data quality is and how to evaluate it. Consequently, practitioners either rely on their own intuition or write custom code to evaluate their data. We discuss potential reasons for this phenomenon and opportunities for alignment.