Back to Details

academic-search

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Academic Search — Paper Discovery

学术搜索——论文发现

Overview

概述

Search academic papers on arXiv using the 
paper-search
CLI tool. No API key required — uses the free arXiv API directly.
使用
paper-search
CLI工具在arXiv上搜索学术论文。无需API密钥——直接调用免费的arXiv API。

When To Use

使用场景

  • User asks to find/search papers on a topic
  • User wants paper recommendations
  • User needs to download a paper PDF
  • User mentions arXiv, academic papers, or research literature
  • 用户要求查找/搜索某一主题的论文
  • 用户想要论文推荐
  • 用户需要下载论文PDF
  • 用户提及arXiv、学术论文或研究文献

Commands

命令

Search papers by keyword

按关键词搜索论文

bash
paper-search search "transformer attention mechanism" --max 5
bash
paper-search search "transformer attention mechanism" --max 5

Search with category filter

按分类筛选搜索

bash
paper-search search "reinforcement learning" --max 10 --categories cs.AI,cs.LG
bash
paper-search search "reinforcement learning" --max 10 --categories cs.AI,cs.LG

Get paper details by arXiv ID

通过arXiv ID获取论文详情

bash
paper-search details 1706.03762
bash
paper-search details 1706.03762

Download paper PDF

下载论文PDF

bash
paper-search download 1706.03762 --output /workspace/papers/
bash
paper-search download 1706.03762 --output /workspace/papers/

JSON output (for programmatic use)

JSON输出（供程序化调用）

bash
paper-search search "neural networks" --max 3 --json
bash
paper-search search "neural networks" --max 3 --json

Example Workflows

示例工作流

Find important papers on a topic

查找某一主题的重要论文

bash
undefined
bash
undefined

Search for recent transformer papers

搜索近期的transformer相关论文

paper-search search "vision transformer ViT" --max 10 --categories cs.CV
paper-search search "vision transformer ViT" --max 10 --categories cs.CV

Get details of the seminal paper

获取开创性论文的详情

paper-search details 2010.11929
undefined
paper-search details 2010.11929
undefined

Build a reading list

构建阅读清单

bash
undefined
bash
undefined

Search and save results

搜索并保存结果

paper-search search "large language models" --max 20 --json > /workspace/papers/llm-papers.json
paper-search search "large language models" --max 20 --json > /workspace/papers/llm-papers.json

Download key papers

下载关键论文

paper-search download 2005.14165 -o /workspace/papers/ # GPT-3 paper-search download 2302.13971 -o /workspace/papers/ # LLaMA
undefined
paper-search download 2005.14165 -o /workspace/papers/ # GPT-3 paper-search download 2302.13971 -o /workspace/papers/ # LLaMA
undefined

Research a specific author

研究特定作者的论文

bash
paper-search search "au:Hinton" --max 10
bash
paper-search search "au:Hinton" --max 10

arXiv Categories

arXiv分类

Common categories for filtering:
CategoryDescription
cs.AIArtificial Intelligence
cs.LGMachine Learning
cs.CLComputation and Language (NLP)
cs.CVComputer Vision
cs.NENeural and Evolutionary Computing
cs.RORobotics
stat.MLStatistics - Machine Learning
math.OCOptimization and Control
常用筛选分类：
分类描述
cs.AI人工智能
cs.LG机器学习
cs.CL计算语言学（NLP）
cs.CV计算机视觉
cs.NE神经与进化计算
cs.RO机器人学
stat.ML统计学-机器学习
math.OC优化与控制

Search Tips

搜索技巧

  • Use quotes for exact phrases: 
    "attention is all you need"
  • Search by author: 
    au:Vaswani
  • Search by title: 
    ti:transformer
  • Combine: 
    "neural network" AND au:LeCun
  • arXiv ID format: 
    1706.03762
    or 
    cs.AI/0001001
  • 使用引号精确匹配短语：
    "attention is all you need"
  • 按作者搜索：
    au:Vaswani
  • 按标题搜索：
    ti:transformer
  • 组合搜索：
    "neural network" AND au:LeCun
  • arXiv ID格式：
    1706.03762
    或 
    cs.AI/0001001

Output Locations

输出位置

  • Papers list: stdout or 
    --json
    to file
  • Downloaded PDFs: 
    /workspace/papers/
    (default)
  • 论文列表：标准输出或通过
    --json
    保存至文件
  • 下载的PDF：默认路径为
    /workspace/papers/

Important Notes

重要说明

  • arXiv API has rate limits; avoid rapid-fire requests
  • PDFs are typically 1-10 MB each
  • Search results are sorted by relevance by default
  • Published dates are in YYYY-MM-DD format
  • arXiv API有调用频率限制；避免频繁连续请求
  • PDF文件大小通常为1-10 MB
  • 搜索结果默认按相关性排序
  • 发布日期格式为YYYY-MM-DD