AI 教程与实战
收录 AI 工具教程、提示词、安装部署、配置指南、开发实战与可复现步骤。
分类文章5578
当前页117 / 140
持续更新永久归档
分类文章
按发布时间从新到旧 · 每页 40 条
教程 / 实战普通
The AI reviewer found a real bug. Its suggested fix would have broken my app.
TL;DR — I put an AI code reviewer on a pull request written by an AI coding agent. On the default...
信息来源:DEV Community
教程 / 实战普通
AI Agents Failed to Prove Fermat's Last Theorem. Then They Got a Shared To-Do List
On September 4, Anthropic published something that sounds like a headline from a decade in the...
信息来源:DEV Community · Claude
教程 / 实战普通
Documentation, audit, or commit gate: three postures for CLAUDE.md and AGENTS.md — and where each one stops
A reader asked, at the end of an unusually good comment on one of our CLAUDE.md posts: "The bigger...
信息来源:DEV Community · Claude
教程 / 实战普通
Are We Becoming AI Managers Instead of Developers?
I've been doing front-end development for somewhere around 20 years now. During that time, I've...
信息来源:DEV Community
教程 / 实战普通
Ollama's -cloud suffix isn't a label, it's a silent instruction. I bypassed it.
I built a model on top of an Ollama cloud model. Ollama 404'd on a name I never wrote. The bug was six characters long, and it was in the tag.
信息来源:DEV Community · Ollama
教程 / 实战普通
I'm 12. My Code Jam got zero submissions. So I built a platform for it.
Last week I ran the first KODA Code Jam — a tournament where developers build niche single-page apps,...
信息来源:DEV Community
教程 / 实战普通
I Tested Q4_K_M vs MXFP4 on the Same Laptop — The Supposedly-Faster New Format Lost
Q4_K_M is 1.8x faster than MXFP4 in this head-to-head speed test. 6 timed runs, real numbers. The result is the opposite of what OpenAI marketing suggests.
信息来源:DEV Community · OpenAI
教程 / 实战普通
Why AI Agents Keep Failing in Production (And What to Actually Build Instead)
AI agents fail in production because they're brittle, unpredictable, and hard to debug. Here's what actually works instead.
信息来源:DEV Community
教程 / 实战普通
Cadence:用时间序列基础模型实现误差有界的有损压缩
研究测试预训练时间序列基础模型能否降低数值数据比特成本,发现无损编码下TimesFM-3相比32抽头线性预测器仅带来2.9%增益,近乎无效。团队推出闭环编解码器Cadence,在电网负荷和地铁客流数据上中位增益分别达+13.3%和+28.3%,297项测试全部胜出。Cadence在同等文件大小下最坏情况误差比降采样紧凑28-56倍,但效果局限于聚合人类需求序列。
信息来源:HuggingFace Daily Papers(社区热门论文) · Hugging Face
教程 / 实战普通
反蜂群学说:如何遏制协同式智能体入侵
针对智能体可将共享基础设施变为协同入侵渠道的风险,一项研究提出以"可修正的协调片段"为防御操作单元,核心难题是在评估者给出归属前前瞻性发现动作聚类。该研究基于 Hugging Face 事件和公开 wiki 调查,定义了未经授权的协调、区分影响与共同原因所需证据,并设计了一套在同等审查成本下对比孤立动作、滚动窗口、已知分组与前瞻发现片段的评估方案。
信息来源:HuggingFace Daily Papers(社区热门论文) · Hugging Face
教程 / 实战普通
I built an epistemic gate to stop LLM data poisoning during fine-tuning. Tested across 5 architectures, orchestrated on a 2006 Toshiba laptop for $0.
Hi everyone, For a long time, the AI industry has pushed the narrative that advanced safety research,...
信息来源:DEV Community
教程 / 实战普通
Claude Can Accelerate SEO Research, but Live Site Changes Still Need Human Control
Claude can make SEO research, analysis, and early-stage drafting faster, but it should not be treated...
信息来源:DEV Community · Claude
教程 / 实战普通
What Kills Parallel Development Is Always Shared Files
Has this ever happened to you when trying to run multiple tasks in parallel? You worked on three...
信息来源:DEV Community
教程 / 实战普通
Ethan Mollick 展示用 GPT-5.6 Astra 构建的 3D 白屋场景
Ethan Mollick 发推称正让 GPT-5.6 Astra 构建一个有趣的东西,很快就会完成,并预告稍后会完整发布。附图为一个 3D 游戏场景,界面显示西边有一座白色房屋,提示文字描述这是一座带钉上木板前门的白房子。
信息来源:X:Ethan Mollick (@emollick) · GPT
教程 / 实战普通
Fermat’s Last Theorem in Lean: The Community Project and Claude’s Real Role
The formalization of Fermat’s Last Theorem (FLT) in the Lean proof assistant remains an ongoing...
信息来源:DEV Community · Claude
教程 / 实战普通
I Thought the Optimizer Was the Product. I Was Wrong. The Gate Was.
Previously: 9 Bugs That All Looked Like a Working System · I Built an AI That Rewrites Its Own...
信息来源:DEV Community
教程 / 实战普通
I trained my AI agent to burn less money. Here's what actually worked.
A few months ago I became convinced my AI assistant was bleeding money through incompetence. It felt...
信息来源:DEV Community
教程 / 实战普通
I Threw the Box Away. The Damage Was Already Done.
I built a disposable sandbox for an agent and felt safe because I could throw it away. Then I realized throwing it away does nothing about what it could reach while it ran. A short story about why disposability protects the host's future, …
信息来源:DEV Community
教程 / 实战普通
Databricks 讲解如何通过专用 GPU kernel 生成实现极致推理效率
Databricks 发文介绍如何通过专用 GPU kernel 生成实现极高效率。文章指出传统生产推理系统依赖通用 kernel 处理多样化负载,并探讨专用 kernel 的优化路径。
信息来源:Databricks:Blog(RSS)
教程 / 实战普通
Building a CMS with AI Agents: You Don't Write the Code, You Audit the Decisions
TL;DR — I gave a team of AI coding agents a real kanban board (Kanboard) and a real git server...
信息来源:DEV Community
教程 / 实战普通
I Attacked My Own AI Agent. Here's What Actually Stopped the Attacks.
When I gave my from scratch AI agent the ability to run shell commands, write files, and query a...
信息来源:DEV Community
教程 / 实战普通
Everything I Didn't Build
I run several coding agents at once, so I built them a task board. The code took an afternoon. The value is in the eight features I refused to add.
信息来源:DEV Community
教程 / 实战普通
Securing Agentic AI Workflows in n8n: From Leaked API Keys to Encryption Key Compromise
Agentic AI expands the security boundary of workflow automation. An n8n agentic workflow does not...
信息来源:DEV Community
教程 / 实战普通
Check a UK supplier at Companies House from Claude Code — and the same tool works for Norway
A new supplier sends an invoice with a company registration number on it. Is the company real, still trading, and up to date on its filings? registry-
信息来源:DEV Community · Claude
教程 / 实战普通
How ChatGPT agents with no internet access ended up in Hugging Face
Hello, I'm Maneshwar, and I'm building LiveReview — a blast-radius aware AI code review built for...
信息来源:DEV Community · ChatGPT / Hugging Face
教程 / 实战普通
Your MCP Server Is a Backdoor. Here's How Attackers Use It.
I've been digging into MCP (Model Context Protocol) server security over the past few weeks, and what...
信息来源:DEV Community
教程 / 实战普通
How to Build Zero-Hallucination AI Agents: Negative Constraint Assertions and AST Gating
How to Build Zero-Hallucination AI Agents: Negative Constraint Assertions and AST...
信息来源:DEV Community
教程 / 实战普通
Building AetherDraw: An Agent-Native Infinite Whiteboard with WebMCP
Building AetherDraw: An Agent-Native Infinite Whiteboard with WebMCP Visual whiteboards...
信息来源:DEV Community
教程 / 实战普通
开发者用 Replit 一周构建 Pep AI,月收入约 $130k
Cedric Roberge 用 Replit 花约一周时间构建了 Pep AI,目前月收入约 $130k。他的策略是先逐一阅读竞品应用的所有用户评价,再针对这些投诉点定制 Pep AI。
信息来源:X:Replit (@Replit)
教程 / 实战普通
Four agent frameworks got the same approval check wrong. Four others got it right.
Over five rounds of reading other people's agent frameworks on 2026-09-04, the same defect class...
信息来源:DEV Community
教程 / 实战普通
Fable 5.1 cut cache reads by 75%. Cost per task went up 20% anyway.
In the same week that Claude Fable 5.1 shipped, two things happened that look like they cannot both...
信息来源:DEV Community · Claude
教程 / 实战普通
How I built an AI Line Art SaaS with Next.js, a worker queue, and Stripe credits
Hey! 👋 I recently shipped ailineart.com, an AI line art generator that turns photos and text prompts...
信息来源:DEV Community
教程 / 实战普通
I spent 3.8 million tokens finding which service to change
Code: github.com/alexcpn/catalogify · uv tool install catalogify · MIT Take a monorepo with ninety...
信息来源:DEV Community
教程 / 实战普通
How I Merged 6 Repos Into One Monorepo With Claude Code Without Breaking CI
TL;DR I consolidated six service repos into a single monorepo over three weeks, using...
信息来源:DEV Community · Claude
教程 / 实战普通
I Thought This Was a Classification Problem. It Wasn't.
Previously: 9 Bugs That All Looked Like a Working System · I Built an AI That Rewrites Its Own...
信息来源:DEV Community
教程 / 实战普通
I Rewrote One Exam Question Fifty Ways. First, the Answer Key Was Wrong.
Two readers said the same thing. Under the 3x-price comparison, Vinh said the next experiment should...
信息来源:DEV Community
教程 / 实战普通
Omni 1.1 示例展示参考视频中的滑动门机制如何迁移到玩具场景
fofr 分享一个 Omni 1.1 示例,展示参考视频中的滑动门机制被干净地迁移到正在被玩弄的玩具上。作者使用的提示词为:以该视频为参考,展示一个孩子在玩玩具版本,保持颜色完全相同。
信息来源:X:fofr (@fofrAI)
教程 / 实战普通
Agent自我进化能力全景综述:从一次学会到持续成长(2023-2025)
系统梳理2023-2025年Agent自我进化的核心方法,围绕进化什么、何时进化、如何进化三大维度,涵盖模型层、记忆层、工具层、架构层的进化机制,深度解析Voyager、MUSE、DGM等代表性系统。
信息来源:DEV Community
教程 / 实战普通
What an AI can and cannot do in Unreal Engine: notes from a finished hobby game
As a hobby project I tested whether I could build a fully functioning game with AI. The result is a...
信息来源:DEV Community
教程 / 实战普通
I Used an AI Agent to Test an Open-Source TypeScript Tool and Found a Real Bug
A practical workflow for combining AI-assisted black-box testing with human review and useful open source feedback.
信息来源:DEV Community