My research group:

Overarching Research Themes

*Extracted by GPT-4, there may be inconsistencies.* #### *Ethics and Human-Centered AI* My research group explores the intersection of ethics and artificial intelligence, critically examining how AI systems can align with human values and social norms. Our work investigates user perceptions and strategies for addressing biases in AI companions, as evidenced in the important paper [User-Driven Value Alignment: Understanding Users' Perceptions and Strategies for Addressing Biased and Discriminatory Statements in AI Companions](https://arxiv.org/abs/2409.00862). We also delve into the complexities of evaluating fairness and accountability in AI systems, highlighted in the paper [Diverse Perspectives on AI: Examining People's Acceptability and Reasoning of Possible AI Use Cases](https://arxiv.org/abs/2502.07287). Additionally, we address safety concerns in human-AI interactions through innovative frameworks, such as demonstrated in [HAICOSYSTEM: An Ecosystem for Sandboxing Safety Risks in Human-AI Interactions](http://arxiv.org/abs/2409.16427). #### *Narrative Analysis and Empathy* My research group explores the nuances of narratives and their impact on human empathy, focusing on how stories can convey complex social dynamics. The paper [HEART-felt Narratives: Tracing Empathy and Narrative Style in Personal Stories with LLMs](https://arxiv.org/abs/2405.17633) showcases our methodology for analyzing the emotional components of narratives captured by AI. We also assess the variability of narrative perceptions in social media through our critical examination in [The Empirical Variability of Narrative Perceptions of Social Media Texts](https://aclanthology.org/2024.emnlp-main.1113/). Our research emphasizes how narratives can influence understanding and empathy across communities, illustrated by [Modeling Empathic Similarity in Personal Narratives](https://arxiv.org/abs/2305.14246). #### *Social Intelligence in AI Interactions* My research group explores the development and evaluation of social intelligence in AI agents, aiming to enhance interactions between humans and machines. A significant contribution is from the study on [Is This the Real Life? Is This Just Fantasy? The Misleading Success of Simulating Social Interactions With LLMs](http://arxiv.org/abs/2403.05020), which critiques the effectiveness of current AI simulations and their ability to replicate human-like interactions. We delve deeper into the nuances of AI responses with our work, [SOTOPIA: Interactive Evaluation for Social Intelligence in Language Agents](https://arxiv.org/abs/2310.11667), which provides frameworks for assessing how well AI agents understand social contexts. Moreover, we investigate the concept of Theory of Mind in AI, laid out in our paper [Clever Hans or Neural Theory of Mind? Stress Testing Social Reasoning in Large Language Models](https://arxiv.org/abs/2305.14763). #### *Mitigating Toxic Language in AI* My research group explores methods for identifying and mitigating toxic language in AI systems and their outputs. Our pivotal study, [Counterspeakers’ Perspectives: Unveiling Barriers and AI Needs in the Fight against Online Hate](https://arxiv.org/abs/2403.00179), addresses the challenges faced in combating online hate speech through AI intervention. We also present a multilingual approach to evaluating AI's toxic responses in [PolygloToxicityPrompts: Multilingual Evaluation of Neural Toxic Degeneration in Large Language Models](https://arxiv.org/abs/2405.09373) which examines biases in language models across cultures. Lastly, our research examines contextual reasoning to enhance AI responses in [COBRA Frames: Contextual Reasoning about Effects and Harms of Offensive Statements](http://arxiv.org/abs/2306.01985), driving forward the conversation on responsibility in AI deployments.