Major: Information Security (undergraduate)

• Awards: Nankai University Academic Excellence Scholarship (2025), Outstanding Key Member of the University Communist Youth League Committee (2022)

Core Team Member

• Conducted security risk assessment for LLMs services, performing deployment research, security testing, and data analysis.
• Investigated localized deployment of 36 model services (Ollama, OpenLLM, AnyLLM), executed 16 unauthenticated vulnerability attacks with comprehensive effect analysis.
• Built a web testing platform using Django, implementing UI design and functional interaction with Python, JSON, and CSS.
• Established automated backend data processing pipelines for scraping, cleaning, and report generation, maintaining data via Navicat-MySQL.
• Achievement: National First Prize, eassy: A First Look at Large Language Model Applications in the Wild from Dual Perspectives-ACM IMC 2025 poster

Team Leader

• Directed a unified model for multi-modal remote sensing object detection project (SM3Det framework).
• Employed Swin-Tiny backbone network and self-supervised pretraining via masked image modeling.
• Constructed a multimodal remote sensing benchmark dataset (SARDet-100K, DOTA, DroneVehicle, Changguang Satellite images).
• Deployed Python core code on lab servers, merging detection module with multimodal task structures.
• Achievements: 19.0% and 15.3% mAP gains on 1.4K and 2.8K image slices respectively. Won third prize in Competition Topic D.

• Solved Problem B "Multi-stage Production Optimization" using Python-based data processing and algorithm design.
• Developed data cleaning, analysis pipelines, and Monte Carlo simulations.
• Visualized results via Matlab and created system architecture flowcharts using Visio and ProcessOn.
• Completed a 24-page modeling paper.

Research Intern

• Conducted research on multimodal remote sensing object detection, optimizing SM3Det with MOE modules and DLA structures.
• Authored an independent Zhihu column "SM3Det: All-in-One Multimodal Remote Sensing Object Detection!".
• Contributed to the lab's AI Agent productization project, advancing "Reduct" app development.
• Serving as project lead for the "Internet + Reduct" initiative.

• Built an encrypted database transformation project using plaintext encryption technology.
• Built system modules with PHP and ODBC interfaces for "zero-code modification" transformation.
• Produced one research paper and a comprehensive project report.