#AI编程
7694 related articles
Programming-Language-Aware Tokenization: The Key to Better Generalization in Code Models
Experiments show that expanding the vocabulary to preserve lexical token boundaries in programming languages improves Qwen3.5-4B from 45.7% to 52.4% on HumanEval and from 39.0% to 47.0% on HumanEval+, with a slight decline on MBPP. Programming-language-aware tokenization reduces the model's learning burden and improves code generalization.
Kimi K3.1 Pro Launches on AI Gateway: Upgraded Coding and Agent Capabilities
Kimi K3.1 Pro and Kimi K3.1 Pro Highspeed launch on AI Gateway with upgraded coding, 3D modeling, financial analysis, and agent capabilities, retaining 1M token context, with weights planned before November 4.
Vercel Adds Build Cache Size Display to Deployments, Making Performance Optimization More Transparent
Vercel's deployment summary now shows build cache size, revealing cache savings and proximity to build machine limits. CLI and API support help developers optimize build performance and cost.
AI SDK Svelte 5.0.151 Released: The Evolution of the AI Toolchain Behind a Minor Version Update
AI SDK Svelte releases version 5.0.151, synchronized with core package ai@7.0.151. A look at this patch update, upgrade advice for Svelte AI developers, and the AI SDK toolchain's maintenance cadence.
AI SDK Sandbox Tool just-bash Releases 1.0.162 Patch Update
AI SDK sandbox tool @ai-sdk/sandbox-just-bash releases patch 1.0.162, synchronized with @ai-sdk/harness. This article explains the version signals, linkage mechanism, and practical impact for developers.
AI SDK Ecosystem Update: sandbox-vercel Releases 1.0.162 Patch
AI SDK ecosystem update: @ai-sdk/sandbox-vercel releases 1.0.162 patch alongside @ai-sdk/harness. We cover the changes, SemVer meaning, and impact for developers.
AI SDK TUI 1.0.152 Patch Released: Terminal AI Development Tools Iterate Again
AI SDK terminal UI component @ai-sdk/tui releases patch version 1.0.152, syncing with core package ai@7.0.151. This article examines the update's positioning, its value for terminal AI development scenarios, and the maintenance status behind the release cadence.
@ai-sdk/tool-bash 1.0.5 Update: Bash Tool Patch for AI SDK
@ai-sdk/tool-bash releases patch 1.0.5, syncing the ai@7.0.151 dependency with two commits. This article analyzes the AI SDK Bash tool package update, its upgrade impact on developers, and the tooling trend.
AI SDK Workflow 2.0.83 Released: Patch Update and Version Dependencies
AI SDK Workflow 2.0.83 patch released, with the core change upgrading the ai package to 7.0.151. This article explains the significance of this minor update, the ecosystem signals behind semantic versioning, and upgrade advice for developers.
@ai-sdk/vue 4.0.151 Patch Release: Vue Integration Stability Update
@ai-sdk/vue 4.0.151 patch released, syncing ai@7.0.151. Learn what's in this Vue AI SDK update, how to upgrade, and what it means for developers.
AI SDK workflow-harness 1.0.162 Patch Release
@ai-sdk/workflow-harness releases patch version 1.0.162, with @ai-sdk/harness updated in sync. This article covers the release overview, patch update implications, and upgrade advice.
Andrew Ng Teams Up with Anthropic: A Complete Walkthrough of the Claude Code Hands-On Tutorial
Andrew Ng and Anthropic launch a Claude Code short course covering clear context, parallel sessions, MCP servers, and three hands-on projects: a RAG chatbot, Jupyter data analysis, and Figma-to-frontend.
JetBrains Incubates ThinkRail: A Pi-Based Agentic IDE with a New Spec-Driven Development Experience
JetBrains Innovation Hub's open-source agentic IDE ThinkRail, built on Pi, offers spec-driven development, Git worktree parallel workspaces, and Monaco editor interaction—completely free. This article details installation, workflow, and ecosystem integration.
Vercel CLI Adds feedback Command: Report Bugs Directly from the Terminal, Usable by AI Coding Agents Too
Vercel CLI adds the vercel feedback command, letting developers submit bugs, friction reports, and feature requests directly from the terminal, with -f structured fields for AI coding agents to report task context.
Vercel Releases @ai-sdk/tool-web-fetch: Web Fetching Tool for AI Apps Lands
Vercel releases @ai-sdk/tool-web-fetch 1.0.0, the AI SDK ecosystem's first official web fetching toolkit. This article covers the initial release, its significance for AI app tool calling chains, and compliance and stability boundaries developers should note when integrating.
Is the Agent Harness Dead? Cursor's Early Engineer Breaks Down the Real Architecture of AI Coding Agents
Cursor early engineer Jediah breaks down the real architecture of the agent harness: why "the harness is dead" is only half right, how the data flywheel lets the harness obsolete itself, and how Grokbot, software factories, and primitive design open new frontiers for agents.
Python Visitor Pattern and Better Alternatives Explained
A deep dive into Python's visitor pattern and four alternatives: method pattern, pattern matching, dictionary mapping, and functools.singledispatch. Choose the best decoupling design by change frequency.
357 Days of Deep Experience: Why Omarchy Made Me Stop Distro-Hopping
A 357-day deep dive into Omarchy: 59-second install, out-of-the-box audio and GPU support, the Quattro QuickShell rewrite migration pains, and why Arch users should try it.
Cursor Can Now Control Your Computer: AI Agents Can Drive Desktop Apps for Testing
Cursor adds desktop control: AI agents can drive desktop apps like the iOS Simulator to test their own features. We analyze the impact on dev workflows, possible technical approaches, and reliability concerns.
Vibecoding a Custom Inference Engine: High Hopes, Harsh Reality
A Reddit developer shares the real experience of vibecoding a custom inference engine: running GGUF Q4 models in the browser with WebGL2 and GLSL, only to fall into tokenizer debugging hell. This article analyzes the engineering challenges of browser-side LLM inference and why cloning llama.cpp is the more rational choice.