Survey data are common and invaluable in social science research for understanding population processes and supporting policymaking and planning. Depending on the nature and scale, survey data sharing comes with privacy risks, and data collectors and agencies are constrained by disclosure permissions, limiting usage across research groups and institutes. Previous methods for synthetic data generation and deidentification may not entirely prevent information disclosures, or they may sacrifice data quality and granularity. Using a large-scale national voter file at both national and state levels, this paper introduces an end-to-end pipeline to streamline synthetic data generation and evaluation for survey researchers. This study selects four generative approaches based on different statistical assumptions: the regression-based Synthpop, the generative deep learning-based CTGAN and TVAE, and the large language model-based REaLDTabFormer, and compares them to the baseline synthetic minority oversampling technique (SMOTE). We consider three key dimensions of evaluation (utility, fidelity, and privacy) to highlight the strengths and weaknesses of each approach, and systematically evaluate across various datasets and training sizes. The results reveal that Synthpop is optimized for general utility (i.e., fidelity), while TVAE excels in downstream applications (i.e., target-specific utility) but compromises on general utility and potentially risks data overfitting. REaLDTabFormer demonstrates a balanced performance in both general and target-specific utility, whereas CTGAN offers the best privacy protection. We recommend that future researchers select a generative method by considering the trade-offs between performance across various evaluation dimensions, training size, data type, and computational infrastructure.
论文检索
输入标题、作者或关键词，从 1,056 篇学术成果中精准定位
To represent the causal relationships between variables, a directed acyclic graph (DAG) is widely utilized in many areas, such as social sciences, epidemics, and genetics. Many causal structure learning approaches are developed to learn the hidden causal structure using deep learning approaches. However, these approaches have a hidden assumption that the causal relationship remains unchanged over time, which may not hold in real life. In this paper, we develop a new framework to model the dynamic causal graph where the causal relations are allowed to be time-varying. We incorporate the basis approximation method into the score-based causal discovery approach to capture the dynamic pattern of causal graphs. Utilizing the autoregressive model structure, we could capture both contemporaneous and time-lagged causal relationships while allowing them to vary with time. We propose an algorithm that could provide both past-time estimates and future-time predictions on the causal graphs, and conduct simulations to demonstrate the usefulness of the proposed method. We also apply the proposed method for the covid-data analysis, and provide causal estimates on how the effect of policy restriction changes.
The online information ecosystem enables influence campaigns of unprecedented scale and impact. We urgently need empirically grounded approaches to counter the growing threat of malicious campaigns, now amplified by generative AI. But, developing defenses in real-world settings is impractical. Social system simulations with agents modelled using Large Language Models (LLMs) are a promising alternative approach and a growing area of research. However, existing simulators lack features needed to capture the complex information-sharing dynamics of platform-based social networks. To bridge this gap, we present SandboxSocial, a new simulator that includes several key innovations, mainly: (1) a virtual social media platform (modelled as Mastodon and mirrored in an actual Mastodon server) that enables a realistic setting in which agents interact; (2) an adapter that uses real-world user data to create more grounded agents and social media content; and (3) multi-modal capabilities that enable our agents to interact using both text and images---just as humans do on social media. We make the simulator more useful to researchers by providing measurement and analysis tools that track simulation dynamics and compute evaluation metrics to compare experimental results.
Recent advancements in AI have allowed for more nuanced agent-based models (ABMs). These simulations offer a non-invasive way to evaluate policies in diverse and complex social contexts. Our project focuses on developing ABMs to examine the impact of legal and social norms on inequity, particularly in scenarios where systemic discrimination affects human well-being. Our research is grounded on the Capability Approach, which provides a comprehensive framework to assess inequity in terms of real opportunities, underpinning the United Nations Sustainable Development Goals (SDGs). By defining representative (i) agent profiles, (ii) agent decision-making, and (iii) agent environment, this work aims to enhance the realism of ABMs and provide valuable insights for policy-making. To achieve this, we are in the process of developing (i) a novel population synthesis method to generate agent profiles that include motivators of behaviour, (ii) a decision-making model based on Markov decision processes (MDPs) that integrates values and needs in short-term and long-term rewards, and (iii) an ABM simulation to assess the impact of norms on inequity in terms of the real opportunities of individuals, among other indicators.
This paper explores AI-mediated human-pet musical co-creation from an interdisciplinary perspective, leveraging recent advancements in animal-assisted therapy. These advancements have shown significant psychosocial benefits, especially in reducing anxiety and enhancing social engagement. Building on these findings, this study innovatively employs pet vocal timbres as 'digital avatars' to enhance emotional investment during the music creation process. We propose PetCoCre, a novel system that applies pet vocal timbres in three distinct character paradigms within AI music creation: (1) PetRhythm: using pet voices as rhythmic percussion through beat synchronization. (2) PetMelody: enabling pet voices to act as melodic instruments via pitch-shifting alignment. (3) PetVocalia: utilizing pet vocal timbres as the target timbre for SVC (Singing Voice Conversion), where the converted singing voice replaces the original singer's voice, thus preserving the original semantic content. Beyond these character paradigms, our technical innovation lies in proposing SaMoye, the first open-source, high-quality zero-shot SVC model that effectively overcomes existing methods' zero-shot limitations by employing mixed speaker embeddings for timbre enhancement and leveraging a large-scale singing voice dataset. In our experiments, we collected dog and cat vocalization data from pet stores and conducted experiments with 30 participants. Results demonstrate that the human-pet co-creation mode led to significant enhancements in pleasure and creative satisfaction compared to solo AI music generation, along with a significant reduction in participants' anxiety levels. Through collaborative art creation, this research pioneers new paradigms for animal-assisted therapeutic interventions and expands the boundaries of AI-assisted creative collaboration.
Probation is a crucial institution in modern criminal law, embodying the principles of fairness and justice while contributing to the harmonious development of society. Despite its importance, the current Intelligent Judicial Assistant System (IJAS) lacks dedicated methods for probation prediction, and research on the underlying factors influencing probation eligibility remains limited. In addition, probation eligibility requires a comprehensive analysis of both criminal circumstances and remorse. Much of the existing research in IJAS relies primarily on data-driven methodologies, which often overlooks the legal logic underpinning judicial decision-making. To address this gap, we propose a novel approach that integrates legal logic into deep learning models for probation prediction, implemented in three distinct stages. First, we construct a specialized probation dataset that includes fact descriptions and probation legal elements (PLEs). Second, we design a distinct probation prediction model named the Multi-Task Dual-Theory Probation Prediction Model (MT-DT), which is grounded in the legal logic of probation and the Dual-Track Theory of Punishment. Finally, our experiments on the probation dataset demonstrate that the MT-DT model outperforms baseline models, and an analysis of the underlying legal logic further validates the effectiveness of the proposed approach.
Minimizing Polarization and Disagreement in the Friedkin–Johnsen Model with Unknown Innate Opinions
PDF ↗The bulk of the literature on opinion optimization in social networks adopts the Friedkin–Johnsen (FJ) opinion dynamics model, in which the innate opinions of all nodes are known: this is an unrealistic assumption. In this paper, we study opinion optimization under the FJ model without the full knowledge of innate opinions. Specifically, we borrow from the literature a series of objective functions, aimed at minimizing polarization and/or disagreement, and we tackle the budgeted optimization problem, where we can query the innate opinions of only a limited number of nodes. Given the complexity of our problem, we propose a framework based on three steps: (1) select the limited number of nodes we query, (2) reconstruct the innate opinions of all nodes based on those queried, and (3) optimize the objective function with the reconstructed opinions. For each step of the framework, we present and systematically evaluate several effective strategies. A key contribution of our work is a rigorous error propagation analysis that quantifies how reconstruction errors in innate opinions impact the quality of the final solutions. Our experiments on various synthetic and real-world datasets show that we can effectively minimize polarization and disagreement even if we have quite limited information about innate opinions.
Responsible AI (RAI) is the science and practice of ensuring the design, development, use, and oversight of AI are socially sustainable---benefiting diverse stakeholders while controlling the risks. Achieving this goal requires active engagement and participation from the broader public. This paper introduces "We are AI: Taking Control of Technology," a public education course that brings the topics of AI and RAI to the general audience in a peer-learning setting. We outline the goals behind the course's development, discuss the multi-year iterative process that shaped its creation, and summarize its content. We also discuss two offerings of "We are AI" to an active and engaged group of librarians and professional staff at New York University, highlighting successes and areas for improvement. The course materials, including a multilingual comic book series by the same name, are publicly available and can be used independently. By sharing our experience in creating and teaching "We are AI", we aim to introduce these resources to the community of AI educators, researchers, and practitioners, supporting their public education efforts.
Responsible AI (RAI) encompasses the science and practice of ensuring that AI design, development, and use are socially sustainable--—maximizing the benefits of technology while mitigating its risks. Industry practitioners play a crucial role in achieving the objectives of RAI, yet there is a persistent a shortage of consolidated educational resources and effective methods for teaching RAI to practitioners. In this paper, we present a stakeholder-first educational approach using interactive case studies to foster organizational and practitioner-level engagement and enhance learning about RAI. We detail our partnership with Meta, a global technology company, to co-develop and deliver RAI workshops to a diverse company audience. Assessment results show that participants found the workshops engaging and reported an improved understanding of RAI principles, along with increased motivation to apply them in their work.
Persuasion is important in numerous situations like healthy habit promotion, and emotional support. As AI gets more involved in our daily life, it becomes critical to study how they can persuade humans and how persuasive they are. In this talk, I will cover (1) how to build such persuasive AI systems that can persuade, negotiate, and cooperate with other humans in the game of Diplomacy. (2) I will also discuss how humans perceive such specialized AI systems. This study validates the necessity of California's Autobot Law and proposes guidance to regulate such systems. (3) As these systems become more powerful, AI safety problems become more important. So I will describe how to persuade AI models to jailbreak them and study AI safety problems. Finally, I will conclude with my long-term vision to further study persuasion from a multi-angle approach that combines Artificial Intelligence, Human-Computer Interaction, and social sciences.
iLLuMinaTE: An LLM-XAI Framework Leveraging Social Science Explanation Theories Towards Actionable Student Performance Feedback
PDF ↗Recent advances in eXplainable AI (XAI) for education have highlighted a critical challenge: ensuring that explanations for state-of-the-art models are understandable for non-technical users such as educators and students. In response, we introduce iLLuMinaTE, a zero-shot, chain-of-prompts LLM-XAI pipeline inspired by Miller (2019)'s cognitive model of explanation. iLLuMinaTE is designed to deliver theory-driven, actionable feedback to students in online courses. iLLuMinaTE navigates three main stages — causal connection, explanation selection, and explanation presentation — with variations drawing from eight social science theories (e.g. Abnormal Conditions, Pearl's Model of Explanation, Necessity and Robustness Selection, Contrastive Explanation). We extensively evaluate 21,915 natural language explanations of iLLuMinaTE extracted from three LLMs (GPT-4o, Gemma2-9B, Llama3-70B), with three different underlying XAI methods (LIME, Counterfactuals, MC-LIME), across students from three diverse online courses. Our evaluation involves analyses of explanation alignment to the social science theory, understandability of the explanation, and a real-world user preference study with 114 university students containing a novel actionability simulation. We find that students prefer iLLuMinaTE explanations over traditional explainers 89.52% of the time. Our work provides a robust, ready-to-use framework for effectively communicating hybrid XAI-driven insights in education, with significant generalization potential for other human-centric fields.
Existing work on the alignment problem has focused mainly on (1) qualitative descriptions of the alignment problem; (2) attempting to align AI actions with human interests by focusing on value specification and learning; and/or (3) focusing on a single agent or on humanity as a monolith. Recent sociotechnical approaches highlight the need to understand complex misalignment among multiple human and AI agents. We address this gap by adapting a computational social science model of human contention to the alignment problem. Our model quantifies misalignment in large, diverse agent groups with potentially conflicting goals across various problem areas. Misalignment scores in our framework depend on the observed agent population, the domain in question, and conflict between agents' weighted preferences. Through simulations, we demonstrate how our model captures intuitive aspects of misalignment across different scenarios. We then apply our model to two case studies, including an autonomous vehicle setting, showcasing its practical utility. Our approach offers enhanced explanatory power for complex sociotechnical environments and could inform the design of more aligned AI systems in real-world applications.
Identifying informative components in binary data is an essential task in many application areas, including life sciences, social sciences, and recommendation systems. Boolean matrix factorization (BMF) is a family of methods that performs this task by factorizing the data into dense factor matrices. In real-world settings, the data is often distributed across stakeholders and required to stay private, prohibiting the straightforward application of BMF. To adapt BMF to this context, we approach the problem from a federated-learning perspective, building on a state-of-the-art continuous binary matrix factorization relaxation to BMF that enables efficient gradient-based optimization. Our approach only needs to share the relaxed component matrices, which are aggregated centrally using a proximal operator that regularizes for binary outcomes. We show the convergence of our federated proximal gradient descent algorithm and provide differential privacy guarantees. Our extensive empirical evaluation shows that our algorithm outperforms, in quality and efficacy, federation schemes of state-of-the-art BMF methods on a diverse set of real-world and synthetic data.
Large language models and other highly capable AI systems ease the burdens of deciding what to say or do, but this very ease can undermine the effectiveness of our actions in social contexts. We explain this apparent tension by introducing the integrative theoretical concept of "mental proof," which occurs when observable actions are used to certify unobservable mental facts. From hiring to dating, mental proofs enable people to credibly communicate values, intentions, states of knowledge, and other private features of their minds to one another in low-trust environments where honesty cannot be easily enforced. Drawing on results from economics, theoretical biology, and computer science, we describe the core theoretical mechanisms that enable people to effect mental proofs. An analysis of these mechanisms clarifies when and how artificial intelligence can make low-trust cooperation harder despite making thinking easier.
We introduce MMVU, a comprehensive expert-level, multi-discipline benchmark for evaluating foundation models in video understanding. MMVU includes 3,000 expert-annotated questions spanning 27 subjects across four core disciplines: Science, Healthcare, Humanities & Social Sciences, and Engineering. Compared to prior benchmarks, MMVU features three key advancements. First, it challenges models to apply domain-specific knowledge and perform expert-level reasoning to analyze specialized-domain videos, moving beyond the basic visual perception typically assessed in current video benchmarks. Second, each example is annotated by human experts from scratch. We implement strict data quality controls to ensure the high quality of the dataset. Finally, each example is enriched with expert-annotated reasoning rationals and relevant domain knowledge, facilitating in-depth analysis. We conduct an extensive evaluation of 36 frontier multimodal foundation models on MMVU. The latest System-2-capable models, o1 and Gemini 2.0 Flash Thinking, achieve the highest performance among the tested models. However, they still fall short of matching human expertise. Through in-depth error analyses and case studies, we offer actionable insights for future advancements in expert-level, knowledge-intensive video understanding for specialized domains.
The Prompt Makes the Person(a): A Systematic Evaluation of Sociodemographic Persona Prompting for Large Language Models
PDF ↗Persona prompting is increasingly used in large language models (LLMs) to simulate views of various sociodemographic groups. However, how a persona prompt is formulated can significantly affect outcomes, raising concerns about the fidelity of such simulations. Using five open-source LLMs, we systematically examine how different persona prompt strategies, specifically role adoption formats and demographic priming strategies, influence LLM simulations across 15 intersectional demographic groups in both open- and closed-ended tasks. Our findings show that LLMs struggle to simulate marginalized groups but that the choice of demographic priming and role adoption strategy significantly impacts their portrayal. Specifically, we find that prompting in an interview-style format and name-based priming can help reduce stereotyping and improve alignment. Surprisingly, smaller models like OLMo-2-7B outperform larger ones such as Llama-3.3-70B.Our findings offer actionable guidance for designing sociodemographic persona prompts in LLM-based simulation studies.
Self-disclosure can provide psychological comfort and social support, but it also carries the risk of unintentionally revealing sensitive information, leading to serious privacy concerns. Research on self-disclosure in Chinese multimodal contexts remains limited, lacking high-quality corpora, analysis, and methods for detection. This work focuses on self-disclosure behaviors on Chinese multimodal social media platforms and constructs a high-quality text-image corpus to address this critical data gap. We systematically analyze the distribution of self-disclosure types, modality preferences, and their relationship with user intent, uncovering expressive patterns unique to the Chinese multimodal context. We also fine-tune five multimodal large language models to enhance self-disclosure detection in multimodal scenarios. Among these models, the Qwen2.5-omni-7B achieved a strong performance, with a partial span F1 score of 88.2%. This study provides a novel research perspective on multimodal self-disclosure in the Chinese context.
Large Language Models (LLMs) demonstrate remarkable ability to comprehend instructions and generate human-like text, enabling sophisticated agent simulation beyond basic behavior replication. However, the potential for creating freely customisable characters remains underexplored. We introduce the Customisable Conversation Agent Framework, which employs LLMs to simulate real-world characters through personalised characteristic feature injection, enabling diverse character creation according to user preferences.We propose the SimsConv dataset, comprising 68 customised characters and 13,971 multi-turn role-playing dialogues across 1,360 real-world scenes. Characters are initially customised using pre-defined elements (career, aspiration, traits, skills), then expanded through personal and social profiles. Building on this, we present SimsChat, a freely customisable role-playing agent incorporating various realistic settings and topic-specified character interactions.Experimental results on both SimsConv and WikiRoleEval datasets demonstrate SimsChat’s superior performance in maintaining character consistency, knowledge accuracy, and appropriate question rejection compared to existing models. Comprehensive ablation studies validate each component’s contribution to overall performance, with the pre-defined aspects framework and scene construction showing particularly significant impact. Our framework provides valuable insights for developing more accurate and customisable human simulacra.Our data and code are publicly available at https://github.com/Bernard-Yang/SimsChat.
Humans have an inherent need for community belongingness. This paper investigates this fundamental social motivation by compiling a large collection of parallel datasets comprising over 7 million posts and comments from Reddit and 200,000 posts and comments from Dread, a dark web discussion forum, covering similar topics. Grounded in five theoretical aspects of the Sense of Community framework, our analysis indicates that users on Dread exhibit a stronger sense of community membership. Our data analysis reveals striking similarities in post content across both platforms, despite the dark web’s restricted accessibility. However, these communities differ significantly in community-level closeness, including member interactions and greeting patterns that influence user retention and dynamics. We publicly release the parallel community datasets for other researchers to examine key differences and explore potential directions for further study.
Warning: This paper contains examples of stereotypes and biases. Large Language Models (LLMs) exhibit considerable social biases, and various studies have tried to evaluate and mitigate these biases accurately. Previous studies use downstream tasks to examine the degree of social biases for evaluation and mitigation. While the output of LLMs highly depends on prompts, prior works evaluating and mitigating bias have often relied on a limited variety of prompts. In this paper, we investigate the sensitivity of LLMs when changing prompt variations (task instruction, few-shot examples, debias-prompt) by analyzing task performance and social bias of LLMs. Our experimental results reveal that LLM rankings fluctuate across prompts for both task performance and social bias. We also confirmed that the impact of format changes can differ for each bias category. Performance improvement from prompt settings may not result in reduced bias. Moreover, the ambiguity of instances is a common factor in LLM sensitivity to prompts across advanced LLMs. We recommend using diverse prompts, as in this study, to compare the effects of prompts on social bias in LLMs.