Zi's Blog
- 2026-05-27_【2026年版】我的AI Agent科研工作流:一个苦涩的教训
- 2026-05-24_当我们与语言模型对话时,我们在与什么对话
- 2026-05-10_Introducing EMatrix — A Matrix-Inspired Emacs Theme
- 2026-05-10_My NixOS Configuration!
- 2026-04-19_对圆融学派的思考与批判
- 2026-04-09_论无我
- 2026-04-02_我的emacs配置介绍-2026版
- 2026-04-01_AI Agent成熟年 我对编程与工作流的思考
- 2026-02-19_OpenClaw 设计细节(三):会话管理、定时任务、节点系统、配置系统
- 2026-02-19_OpenClaw 代码乱读笔记 2
- 2026-02-19_OpenClaw 代码乱读笔记1
- 2026-02-03_我的全新TODO管理系统
- 2025-12-09_对2025年的自我反思以及最后一个月的生存计划
- 2025-09-28_我的创业与开源项目
- 2025-09-28_我的文学创作:自述
- 2025-09-03_Slurm's Configuration and Usage
- 2025-06-03_Best Practice of Crafting Prompts
- 2025-06-02_LLM operates Computers: An Introduction and Framework of RL-driven Agent
- 2025-04-23_My Attempt of a Potentially New TODO Workflow: How to sync Mobile Phone with My Computer.
- 2025-04-07_How to Improve the Impact of My Research
- 2025-02-20_The Hidden Risks of Large Reasoning Models: A Safety Assessment of R1
- 2025-02-17_礼记章句(四)–乐记篇笔记
- 2025-02-17_礼记章句(三)学记篇笔记
- 2025-02-17_礼记章句–少仪篇笔记
- 2025-02-16_Thoughts of Deepseek R1
- 2025-01-05_粤语非系统学习-1
- 2024-12-08_对过程导向与结果导向的重新思考
- 2024-11-09_船山遗书册四《礼记章句上》读后感
- 2024-11-05_Linux: mount the disk into
/home - 2024-10-29_VPN Configuration to Bypass Some Firewalls.
- 2024-10-13_矩阵求导的一般规律
- 2024-09-07_理解Kernel Trick、Gaussian Process以及Neural Tangent Kernel
- 2024-09-01_从川端康成到中上健次:日本纯文学是如何死掉的?在现代社会中重新建构东方美学是可能的吗?
- 2024-09-01_对论文写作的一些评价和方法论
- 2024-04-19_Six principles in one-one meeting with my supervisor
- 2024-03-26_A standard evaluation bench for NLG evaluations
- 2024-03-12_Vocabulary for Academic Writing
- 2024-03-10_Attacks and Defenses (Watermarking) in Large Language Models
- 2024-03-09_中文预训练模型汇总
- 2024-03-09_在GPU上进行训练的一些指南
- 2024-03-09_search on graph系列算法笔记
- 2024-03-09_org mode 简略指南(个人翻译版)
- 2024-03-09_对话系统中的伦理问题小考
- 2024-03-09_/存在与时间/ 引题笔记
- 2023-12-30_Interesting Researches of AI security in CCS 2023
- 2023-12-14_USENIX-23-Fall Interesting Papers
- 2023-11-04_The knowledges behind Crypten
- 2023-10-20_我一生的经历 梁洪盘
- 2023-10-06_Common Lisp notes and thinkings (Part I: Functions)
- 2023-09-22_IELTS Writing Tracks – from the video of Simon
- 2023-09-16_Rethinking the Methodology of Handling EXAMS : a Case Study and Conclusion from IELTs
- 2023-08-15_论南怀瑾等流
- 2023-08-13_对佛教唯识学的思考与分析【废弃】
- 2023-08-12_Access state-of-the-art Large Language Models & APIs
- 2023-08-06_Vocabulary of listening in IELTs from 17 to 11
- 2023-06-16_《瑜伽师地论》(《十七地论》)本地分枝络小结
- 2023-06-06_对《十七地论》中“菩萨地”理念的分析
- 2023-05-27_学会聚焦——良好心态的养成手册
- 2023-05-18_对佛教中“声闻地”理念的批判分析与解读
- 2023-03-30_读恩格斯《家庭、私有制与国家的起源》有感
- 2023-03-09_论《费曼学习法》与个人哲学体系建立
- 2023-02-28_Attacks and Defenses of LLMs: 大型语言模型的攻击与防御
- 2023-02-03_Delphi源码学习与分析
- 2023-02-03_[短]cargo 使用基础
- 2023-02-03_多方安全计算原理与相关库介绍
- 2023-01-20_三水西协议——一种新型的文艺作品创作协议
- 2023-01-20_浅谈GUI库使用要点和递进层次
- 2023-01-20_学习一门新的编程语言的方法和原则
- 2023-01-02_密码学知识思考一之有限域上的计算
- 2022-12-15_新冠阳性症状记录
- 2022-12-08_多方安全计算整理笔记
- 2022-11-21_PhD notes
- 2022-11-06_对个体成长类网络文学的浅显思考
- 2022-05-28_论《矛盾论》与《实践论》
- 2022-05-20_《周易》与算卦
- 2022-05-09_智源社区参与编辑文章总结(不定期更新)
- 2022-03-31_运维管理笔记二: 用户管理与硬盘挂载
- 2022-03-30_为一台全新的服务器配置Nvidia GPU环境!
- 2022-03-21_《资本论》笔记
- 2022-02-11_科目三(费县327国道车管所西)学习笔记
- 2022-01-24_我的科研实验工作流
- 2022-01-16_任务型对话系统中Neural Pipeline系列源码剖析
- 2022-01-09_阅读、理解、和解析MultiWoz数据集
- 2022-01-07_半自动手残拯救器:PC端微信跳一跳的解决方案
- 2022-01-03_我的emacs配置文档介绍
- 2021-12-14_宣言
- 2021-12-08_一篇论文是如何被拒掉的?
- 2021-12-06_对话系统中的安全性问题——一份不完整总结
- 2021-12-06_python进阶使用
- 2021-11-20_【低质量】rust是如何适配并发的?
- 2021-11-19_rust 生命周期(lifetime)解析
- 2021-11-18_为什么rust是安全的?从泛型的角度进一步解析
- 2021-11-17_学习rust之所有权的学习
- 2021-11-17_Zi's Blog
- 2021-11-17_关于huggingface transformers NLG模块实现的基本思路
- 2021-09-21_阅读《Pro Git》有感
- 2021-09-18_FATE: 一种联邦学习框架
- 2021-09-10_Zi Liang (Research Page)
- 2021-09-10_我看斯诺登
- 2021-09-10_写在二十三岁开头的诗
- 2021-09-10_我的学术学习工作流
- 2021-09-10_撰写学术论文的要求
- 2021-09-10_废石居士自传
- 2021-08-09_浅论论文和ppt的形式封装
- 2021-08-09_introduction怎么写?
- 2021-04-06_[长]知识图谱入坑笔记
- 2021-01-01_知识图谱增强的预训练模型
- 2020-07-07_emacs lisp基本
- blogs
- 2026-05-27_LoRA的安全性:轻量微调的"阿喀琉斯之踵"?
- 2026-05-27_Argus:多Agent协作的安全漏洞检测
- 2026-05-27_我的研究之旅:从理解模型到发现漏洞,再到构建防御
- 2026-05-27_The Matthew Effect: AI Coding Tools Make the Rich Richer
- 2026-05-27_DPS: Drawing a Decision Boundary for Large Language Models
- 2026-05-27_MERGE:快速私有文本生成——用密码学保护AI隐私
- 2026-05-27_TEMP: Alignment Signals Are in the Data, Not in RLHF
- 2026-05-27_LoRD:怎么"偷"走一个大语言模型?
- 2026-05-27_Argus: Multi-Agent Ensemble for Full-Chain Security Vulnerability Detection
- 2026-05-27_VIA:当毒化攻击遇上合成数据——一场"病毒感染"
- 2026-05-27_MERGE: Fast Private Text Generation—Protecting AI Privacy with Cryptography
- 2026-05-27_LoRA Security: The "Achilles' Heel" of Parameter-Efficient Fine-Tuning?
- 2026-05-27_Web Memorization:LLM记住了互联网的什么?
- 2026-05-27_ArxivRollBench:你的模型在"作弊"吗?
- 2026-05-27_TEMP:对齐信号不在RLHF里,在语料里
- 2026-05-27_Prompt的措辞敏感性:换一个词,结果天差地别
- 2026-05-27_Web Memorization: What Did the LLM Absorb from the Internet?
- 2026-05-27_The Information Bottleneck of Vision Tokens: How Much Can One Token Hold?
- 2026-05-27_ArxivRollBench: Is Your Model Cheating on the Test?
- 2026-05-27_Mobius Injection: Can a Single Message Paralyze AI Infrastructure?
- 2026-05-27_Matthew Effect:AI编程工具的"马太效应"
- 2026-05-27_DPS:给大模型画一条决策边界
- 2026-05-27_Mobius Injection:一条消息瘫痪AI基础设施?
- 2026-05-27_Prompt Lexical Sensitivity: Change One Word, Get a Different World
- 2026-05-27_LoRD: How to Steal a Large Language Model
- 2026-05-27_My Research Journey: Understanding Models, Uncovering Vulnerabilities, and Building Defenses
- 2026-05-27_Vision Token的信息瓶颈:一个Token到底能装多少信息?
- 2026-05-27_Prompt Leakage:你的系统提示词,藏不住的
- 2026-05-27_Prompt Leakage: Your System Prompts Cannot Hide
- 2026-05-27_VIA: When Poisoning Meets Synthetic Data—A "Virus Infection"
- Poems
- 2022-05-16_感冒
- 2022-05-16_闻山中樵夫作老树歌
- 2022-05-16_春
- 2022-05-16_黄昏雨 赠郭亚冲
- 2022-05-16_思念
- 2021-12-17_所有作品之版权要求
- 2021-12-08_宅时
- 2021-12-08_黑鸟
- 2021-12-08_七月两篇
- 2021-12-08_迷路的人
- 2021-12-08_十一月十一日
- 2021-12-08_雨山
- 2021-12-08_写给桃子的一首诗
- 2021-12-08_狮子狗
- 2021-12-08_我讨厌文学
- 2021-12-08_五月雷雨
- 2021-12-08_偷吃灵药的人
- 2021-12-08_跪梦集:现代诗选编
- 2021-12-08_<跪月集>——古诗词相关