Notes笔记
Markdown-first writing以 Markdown 为先的写作
Draft essays, project logs, and technical notes that connect active paper work, portfolio demos, and long-term research identity.这里放草稿文章、项目日志和技术笔记，用来连接当前论文工作、作品集 Demo 和长期研究身份。
Article文章
Understanding AI Agents in Depth: From Principles to Engineering Practice and Case Studies深入理解AI Agent：从原理到工程实践及案例分析
An in-depth guide to AI agent principles, context, memory, tools, harness engineering, evaluation, multi-agent systems, and representative engineering case studies.从 Agent 基本范式、Context、Memory、Tools 与 Harness Engineering 出发，系统讨论评估、多 Agent 协作及 OpenClaw、Hermes、Codex、Claude Code 的工程实践。
Draft草稿
From Optoelectronics to Computational Imaging and Multimodal AI从光电信息到计算成像与多模态 AI
A draft research-identity essay about moving from optoelectronics toward AI systems for multimodal visual perception, computational imaging, and deployable visual restoration.一篇研究身份草稿：记录我如何从光电信息走向面向多模态视觉感知、计算成像和可部署视觉恢复的 AI 系统。
Draft草稿
RobustIRFusion: Why Fusion Needs Robustness, Not Only Better MetricsRobustIRFusion：为什么融合需要鲁棒性，而不只是更好的指标
A draft project note framing infrared-visible fusion as a robustness problem under misalignment and modality reliability uncertainty.一篇项目笔记草稿：把红外-可见光融合放在错位和模态可靠性不确定下的鲁棒性问题中理解。
Draft草稿
Photo Moire Removal: An AI Camera Restoration DemoPhoto Moire Removal：一个 AI 相机图像恢复 Demo
A draft project page for a small photo demoireing portfolio demo under the broader AI camera restoration line.一个小型拍屏去摩尔纹作品集 Demo 草稿，属于更长期的 AI 相机图像恢复方向。
Report专题报告
AI Agents: A Comprehensive Report智能体（Agent）专题报告
A comprehensive report on what AI agents are, what they can do, and how to use, evaluate, and govern them as verifiable task systems.一份围绕智能体是什么、能做什么以及如何使用、评测与治理智能体的中文专题报告，重点讨论从模型能力走向可委托、可验证、可治理的任务系统。