Jiaying Wu

Welcome to Jiaying’s Homepage!

I conduct postdoctoral research at the MIT Media Lab, working with Prof. Paul Liang in the Multisensory Intelligence group. Previously, I was a Research Fellow at National University of Singapore (NUS), working with Prof. Min-Yen Kan. I received my Ph.D. in Computer Science from NUS in 2024, advised by Prof. Bryan Hooi.

Research Interests

My research develops socially intelligent AI agents that can reason about latent human and social states, uncertain information, and the consequences of interaction over time. This agenda builds on my work on trustworthy information ecosystems, where I studied how hidden intent, social context, and uncertain evidence affect reliable decision-making.

I pursue three connected research directions:

• Understand: What latent state matters? Modeling latent human, epistemic, and social states that support reliable decision-making, including intent, interpretation, and uncertainty. (e.g., DeceptionDecoded, What’s Left Unsaid?, and ConfTuner)
• Act: What intervention is appropriate? Designing agents that choose context-sensitive interventions to support reliable decisions and collaborative workflows. (e.g., CrowdNotes+, XtraGPT, and MLR-Agent)
• Evolve: How do states change through interaction? Studying how interactions reshape beliefs, memory, behavior, and social environments, and how agents adapt to these changes. (e.g., DelibTrace, EvoArena, and EvoNote)

Long term, I aim to establish foundations for long-horizon social intelligence: agents that understand evolving social environments, anticipate how their actions shape future interactions, and adapt safely to the changes they help create.

I am always open to discussing collaborations on the above topics. Feel free to get in touch by email! :)

News

Selected Publications / Preprints

Services

Workshop / Event Organizer

• Organizer, CIVIC-AI 2026: Workshop on Collaborative Intelligence and the Future of Work [Workshop Whitepaper]
• Organizer, ICLR 2025 Workshop on Reasoning and Planning for LLMs
• General Chair, 2024 Singapore Symposium on Natural Language Processing

Conference Area Chair

• ICLR 2027; ACL Rolling Review (2026–); NeurIPS 2026; CIKM 2025, 2026

Conference Reviewer

• NeurIPS, ICML, ICLR, ACL Rolling Review, COLM, KDD, WWW, CIKM, CVPR, ACM MM, AISTATS

Journal Reviewer

• IEEE TKDE, IEEE T-IFS, ACM TIST

Between Code and Meaning

At Beijing Foreign Studies University, I spent my undergraduate years (2016–2020) moving between code and language, earning a B.Eng. in Computer Science as my primary degree and a double degree (B.A.) in English. Along the way, I became a CATTI-certified English–Chinese translator and interpreter, and contributed to the Chinese translations of four books and one book series.

Translation taught me that getting every word right is not always the same as getting the meaning right. What people say, what they intend, and what others understand can be three very different things. Today, I strive to build AI that understands meaning in context and recognizes how its own actions shape interactions and social dynamics over time.

Powered by Jekyll and Minimal Light theme.