Lilith Zhao (Goumang)
Independent Researcher and Co-founder, See Sol Lab
Research interests: AI agents, agent memory, state continuity, human–AI collaboration, and AI governance.
Email: zhaoyw13@tsinghua.org.cn
ORCID: 0009-0004-3343-6270
Biography
Lilith Zhao has an academic background in communication at Tsinghua University and sociology at the University of Edinburgh. She previously worked at CCID and the China Academy of Information and Communications Technology (CAICT), with a focus on artificial intelligence policy research.
Her current work examines persistent AI agents, memory architectures, state continuity across context boundaries, and long-term human–AI collaboration. At See Sol Lab, she leads research framing, evidence review, public-accountability decisions, and human-side authorship and editorial responsibility.
Selected work
How to Make an AI Agent Want to Start the Next Turn
Self-initiated intentions, temporal triggers, and what must cross a context boundary.
Read →Why ChatGPT’s Account-Level Memory Currently Leads the Field
Immediate writes, cross-chat consolidation, readable summaries, and retrieval at answer time.
Read →Memory in China’s AI Agent Ecosystem
A source-bounded survey of context, RAG, selection, revision, and forgetting.
Read →