Undergraduate Student
Science and Technology College of Nanchang Hangkong University, China
Email used for academic correspondence and OpenReview registration: 15015091629@163.com
I am 郭家吉 (Jiaji Guo), an undergraduate student interested in building reliable and auditable AI systems, especially long-lived AI agents that must reason over changing information and make decisions under explicit evidence and safety constraints.
- Institution: Science and Technology College of Nanchang Hangkong University, China
- Status: Undergraduate Student
- Enrollment: 2024–present
- Academic email / correspondence: 15015091629@163.com
- AI Agents
- Large Language Models
- Agent Memory and Long-Term Context
- Evidence-Qualified Decision Making
- Retrieval-Augmented Generation
- AI Safety and Evaluation
- Revision-Aware and Traceable AI Systems
Evidence-Qualified and Revision-Aware Decision Intelligence for Long-Lived AI Agents
Omni-Context explores how long-lived AI agents can:
- determine whether retrieved information is still valid and relevant;
- decide whether available evidence is sufficient before acting;
- respect constraints, authority, and approval boundaries;
- revise, reverse, or invalidate earlier decisions when evidence or outcomes change; and
- remain auditable through explicit evidence and decision lineage.
Repository: https://github.com/guo6x/Omni-context
Name: 郭家吉 · Jiaji Guo
Email: 15015091629@163.com
Affiliation: Science and Technology College of Nanchang Hangkong University, China
Current role: Undergraduate Student
Location: China
This GitHub profile is maintained as my public academic and technical identity page. The name, current affiliation, and contact email shown here correspond to the information I use for academic profiles and research-related communication.