Research
Delta-I Lab
PI：Tonghan WANG
Research Direction: Next-generation foundation models, AI agents, and their economies
Research Group Introduction
Our research focuses on next-generation foundation models, AI agents, and their economies. Our key research areas include:
1. Foundational Breakthroughs in LLMs. We focus on expanding the frontiers of intelligence in LLMs. We are pretraining LLMs based on novel architectures, with the goal of developing LLMs that support unbounded context windows, continuously improve themselves, and dynamically adapt their architectures after deployment.
2. Agents and Agent Economies. We study how intelligence emerges and evolves in individual agents and multi-agent systems through learning and interaction. We also explore and shape societal structures in which humans and intelligent agents coexist, together with the economic foundations that support this coexistence.
Representative Publications
1. “BundleFlow: Deep Menus for Combinatorial Auctions by Diffusion-Based Optimization,” Tonghan Wang, Yanchen Jiang, David C. Parkes. NeurIPS 2025.
2. “On Diffusion Models for Multi-Agent Partial Observability: Shared Attractors, Error Bounds, and Composite Flow,” Tonghan Wang*, Heng Dong*, Yanchen Jiang, David C. Parkes, Milind Tambe. AAMAS 2025 Oral.
3. “GemNet: Menu-Based, Strategy-Proof Multi-Bidder Auctions Through Deep Learning,” Tonghan Wang*, Yanchen Jiang*, David C. Parkes. ACM EC 2024.
4. “Deep Contract Design via Discontinuous Networks,” Tonghan Wang, Paul Dütting, Dmitry Ivanov, Inbal Talgam-Cohen, David C. Parkes. NeurIPS 2023.
5. “Low-Rank Modular Reinforcement Learning via Muscle Synergy,” Heng Dong*, Tonghan Wang*, Jiayuan Liu, Chongjie Zhang. NeurIPS 2022.
6. “Non-Linear Coordination Graphs,” Yipeng Kang*, Tonghan Wang*, Qianlan Yang*, Xiaoran Wu, Chongjie Zhang. NeurIPS 2022 Spotlight.
7. “RODE: Learning Roles to Decompose Multi-Agent Tasks,” Tonghan Wang, Tarun Gupta, Anuj Mahajan, Bei Peng, Shimon Whiteson, Chongjie Zhang. ICLR 2021.
8. “ROMA: Multi-Agent Reinforcement Learning with Emergent Roles,” Tonghan Wang, Heng Dong, Victor Lesser, Chongjie Zhang. ICML 2020.
Group Culture & Mentorship Philosophy
Be bold to seek, to seek the unknown, to follow it to the furthest edge.
Contact information
We are continually seeking outstanding Ph.D. students and research interns who are self-driven and genuinely passionate about research. If our research directions align with your interests, please contact me at twang1@g.harvard.edu.Including a CV, a transcript, a brief description of a research project or problem you have worked on, and a short note explaining which of our research problems interest you the most and why will help me know you better.