Artificial Intelligence Researcher

Yong Dai

Language models, multimodal intelligence, autonomous agents, and embodied intelligence.

R01Foundation Models

R02Multimodal Intelligence

R03Agentic Reasoning

R04Embodied Intelligence

01

About

My research bridges large language models and the complex, multimodal world we live in. I completed my Ph.D. at the University of Electronic Science and Technology of China under the guidance of Professor Zenglin Xu, with research experience at Microsoft and Tencent AI Lab.

I now focus on multimodality, web agents, and embodied intelligence. I am interested in unified models that can perceive, reason, act, and learn through interaction, with the long-term goal of connecting world models and autonomous agents into capable general-purpose intelligent systems.

02

Experience

2021 - 2024

Research Intern and Researcher
Tencent AI Lab

2020 - 2021

Visiting Student
Westlake University

2019 - 2020

Research Intern
Microsoft STCA NLP Group

2018 - 2019

Project Lead
Research collaboration with Nuance

03

Research

Four connected research areas converging toward interactive world models
Four connected research directions converge toward interactive world models.

04

News

Optimizing Agentic Reasoning with Retrieval via Synthetic Semantic Information Gain Reward accepted by ICML 2026.

Web-CogReasoner, a multimodal cognitive reasoning framework for web agents, accepted by ICLR 2026.

E-ViC, which studies embodied visual chains for spatial intelligence, accepted by ACL 2026.

Bridging VLMs and Embodied Intelligence with Deliberate Practice Policy Optimization accepted by ACM MM 2026.

Distribution-Aligned Decoding and InfMasking accepted by NeurIPS 2025.

Nexus and MME-Finance accepted by ACM MM 2025.