Research Interests
- Agent Memory: How AI agents represent, store, retrieve, and update long-term memory to enable continuous learning and reasoning.
- Agent Harness and Self-Evolution: How agent orchestration (harness) coordinates planning, tools, and memory, and enables agents to iteratively improve through feedback and self-generated signals.
- Multi-Agent and Multi-Modal Systems: AI systems that integrate diverse agents and data modalities (e.g., text, image, audio, video) to collaboratively solve complex tasks.
- Agent Applications: Deploying agent systems in real-world settings—especially scientific domains—where they can perform complex reasoning, automate workflows, and generate actionable insights from real data.
Selected Publications
Zefeng He, Siyuan Huang, Xiaoye Qu, Yafu Li, Tong Zhu, Yu Cheng, Yang Yang. “GEMS: Agent-Native Multimodal Generation with Memory and Skills” arXiv preprint arXiv:2603.28088, 2026.
Yang Yang, Takashi Nishikawa, Adilson E. Motter. “Small vulnerable sets determine large network cascades in power grids” Science, 358(6365), eaan3184, 2017.
Yang Yang, Adilson E. Motter. “Cascading failures as continuous phase-space transitions” Physical Review Letters, 119(24), 248302, 2017.
Yang Yang, Takashi Nishikawa, Adilson E. Motter. “Vulnerability and cosusceptibility determine the size of network cascades” Physical Review Letters, 118(4), 048301, 2017.
Adilson E. Motter, Yang Yang. “The unfolding and control of network cascades” Physics Today, 70(1), 32–39, 2017.
For a complete list of publications, see my Google Scholar profile.