Xisen Wang
Xisen Wang
I like to build self-evolving, autonomous systems that can scale. In particular, I study the science of networks of agents for how they should organise better.
I like to build self-evolving, autonomous systems that can scale. In particular, I study the science of networks of agents for how they should organise better.
I'm a PhD student at the Columbia DAP Lab, co-advised by Prof. Eugene Wu and Prof. Kostis Kaffes. Previously, I did my master's thesis with the Torr Vision Group at Oxford, advised by Jindong Gu and Philip Torr.
Google Scholar · GitHub · LinkedIn · xisen.wang@columbia.edu
Publications
How Agents Should Organise to Work Better
- Position: Stop Hardcoding Multi-Agent Workflows That General Agents Will OutgrowXisen WangICML Workshop on Agents in the Wild, 2026
- Revisiting Boids for Emergent Intelligence via Multi-Agent Collaborative Tool-BuildingXisen Wang, Qi ZhangWorkshop on Scaling Environments for Agents
- Startup Success Forecasting Framework (SSFF): A Multi-Agent Framework for Startup Success PredictionXisen Wang, Yigit Ihlamur, Fuat AlicanNeurIPS Workshop on Generative AI in Finance, 2025 · Best Poster Award
How Agents Should Organise to Do Research
- Can AI Scientists Coordinate at Runtime?Zijian Liu, Yangzhixin Luo, Junyu Lu, Yi Li, Yu Chen, David Xu, William F. Shen, Xinchi Qiu, Xisen WangarXiv preprint, 2026
- R&D-Agent-Quant: A Multi-Agent Framework for Data-Centric Factors and Model Joint OptimizationYuante Li, Xu Yang, Xiao Yang, Xisen Wang, Weiqing Liu, Jiang BianNeurIPS, 2025
- Agent for Machine Learning: A Text-to-Model (T2M) ApproachXisen Wang, Jingbo ZhaoFuture of Information and Communication Conference (FICC), 2025
How Agents Should Organise Safely
- How Should Your Agent Talk to Mine? Measuring the Utility–Security Frontier of Cross-Boundary Agentic DelegationXisen Wang, Adel Bibi, Kevin Qinghong Lin, Philip Torr, Jindong GuICML Workshop on Agents in the Wild, 2026
- Distributed Legal Infrastructure for a Trustworthy Agentic WebTomer Jordi Chaffer, Victor Jiawei Zhang, Sante Dino Facchini, Botao Amber Hu, Helena Rong, Zihan Guo, Xisen Wang, Carlos Santana, Giovanni De GasperisarXiv preprint, 2026
Earlier Work
- Data-centric Strategies for Carbon-efficient Carbon-intensity ForecastingXisen Wang, Noa ZilbermanSpringer, 2025
- White Noise Tests on the LSTM Model Trained with Double PendulumXisen WangJournal of Physics: Conference Series, 2021