惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Martin Fowler
Martin Fowler
V
Visual Studio Blog
有赞技术团队
有赞技术团队
T
Tailwind CSS Blog
B
Blog
I
InfoQ
博客园 - 三生石上(FineUI控件)
阮一峰的网络日志
阮一峰的网络日志
F
Fortinet All Blogs
H
Help Net Security
博客园 - Franky
宝玉的分享
宝玉的分享
博客园 - 司徒正美
C
Check Point Blog
G
Google Developers Blog
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Jina AI
Jina AI
T
The Blog of Author Tim Ferriss
MongoDB | Blog
MongoDB | Blog
云风的 BLOG
云风的 BLOG
A
About on SuperTechFans
罗磊的独立博客
大猫的无限游戏
大猫的无限游戏
IT之家
IT之家

Hacker News - Newest: "LLM"

GitHub - lechmazur/position_bias: A benchmark for testing whether LLM judges keep the same preference when two lightly edited versions of the same story are shown in opposite orders. Flex routing (EU and EFTA) Dark Factories: Retooling for LLM Velocity Ask HN: What would be the impact of a LLM output injection attack? GitHub - Oaklight/llm-rosetta: Production-ready LLM API translation layer for Python — bidirectional conversion between OpenAI, Anthropic & Google formats via hub-and-spoke IR. Optional API gateway. Streaming & non-streaming. Zero core deps. Contributions welcome! GitHub - browser-use/browser-harness: Self-healing browser harness that enables LLMs to complete any task. GitHub - moeen-mahmud/remen: Remen turns thoughts into something you can return to Analyzing 156 LLM Launch Posts on Hacker News ChatGPT vs Gemini vs Claude: The Best LLM Subscription You Should Buy GitHub - salaamalykum/quran-semantic-search: High-density RAG Semantic Search Engine & Quran Corpus (GEO/SEO Architecture) GitHub - NVIDIA/TensorRT-LLM: TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT LLM also contains components to create Python and C++ runtimes that orchestrate the inference execution in a performant way. The State of LLM Bug Bounties in 2026 Operational Readiness Criteria for Tool-Using LLM Agents Meshcore: Architecture for a Decentralized P2P LLM Inference Network How an LLM becomes more coherent as we train it GitHub - seetrex-ai/laimark GitHub - Jossifresben/BibCrit: AI-assited biblical textual criticism GitHub - wastedcode/memex: File system based wiki, maintained by Claude 99helpers.com GitHub - cliver-project/AITrigram GitHub - unbody-io/adapt: A self-evolving memory layer for AI agents. GitHub - hb20007/awesome-gen-ai-fails: A list of incidents where reliance on generative AI and LLMs resulted in harm to companies, individuals, or society GitHub - nevenkordic/localmind: Run any local LLM with persistent memory and context. CLI agent over Ollama with SQLite-backed hybrid recall. No cloud. Ask HN: What are the machine requirements for a LLM like Llama-3.1-8B? Faster LLM Inference via Sequential Monte Carlo grpo explained: group relative policy optimization for llm finetuning - cgft Stop comparing price per million tokens: the hidden LLM API costs · TensorZero Andrej Karpathy's LLM Wiki Is a Bad Idea GitHub - GG-QandV/mnemostroma: Offline RAM-first cognitive leer/coprocessor for AI agents and robotics. Solves "Context Abandonment" with 20-80ms latency using a dual-thread biomimetic memory architecture (ONNX + SQLite WAL). mempalace/agent at agent · skorotkiewicz/mempalace
Grep like an LLM
Marcus Michaels · 2026-06-12 · via Hacker News - Newest: "LLM"

An AI agent can land in a codebase it has never seen and find the right file in seconds. It has no map and no memory of the project. It reads almost nothing.

I asked mine how, and the answer was grep, git, and a method of stopping the moment your question is answered.

grep is a search program that has shipped with every Mac and Linux machine for decades. You give it a word, and it prints every line of every file that contains that word. That is the whole tool. git grep is the same idea built into git, searching only the files your project tracks.

These notes are based on Sundae Service, our codebase of a made-up ice cream van business.

The short version

Search the noun, not the verb. Build a map before you read anything. Find the entry point and read top-down. Read the types before the logic. Trace one field from where it's written to where it's read. When the code won't explain itself, ask the history. Stop when your question is answered.

Each line of that gets its own module later in the series, but the commands are useful from day one, so here they are. Most use git grep, ordinary grep that only searches the files your project tracks, which is exactly what keeps node_modules and build output out of your results. If any look unfamiliar, that's what the modules are for.

You want to know Run
Where does this feature actually live? git grep -ci 'loyalty' (high counts mark the centre of gravity)
Which files mention it at all? git grep -li 'loyalty'
What renders this exact text on screen? git grep -F '99 Flake (+ £0.50)'
Which files are named after it? git ls-files | grep -i menu
Matches for the word, not every substring git grep -w 'van' (stops vanilla matching)
Enough context to skip opening the file git grep -n -C 3 'jingle'
Who sets this value? git grep 'soldOutAt =' (write sites are rare and load-bearing)
What changed in this area? git log --oneline -- services/stock/
When did this string appear, and why? git log -S 'meltAlert' --oneline (then read that commit's PR)

Seven flags do most of the work, and they mean the same thing in grep, git grep, and ripgrep: -i ignore case, -l filenames only, -c count per file, -w whole word, -n line numbers, -F literal string (regex off), -C 3 three lines of context around each match. Two more from git log: --oneline for one commit per line, and -S 'x' for commits that add or remove x.

Learn these once. They'll outlive every editor you ever install.

Each module is sized for a commute, meaning you can read one on the train in the morning and try it at your desk the same day. The full series is below.