AI Knowledge Base

Total 1235 articles posts
基于 CrewAI 构建 Multi Agent 系统指南

Guide to Building Multi Agent Systems Based on CrewAI

1. Introduction In the field of Artificial Intelligence (AI), Multi Agent systems are becoming a key technology for solving complex problems and realizing efficient collaboration.CrewAI, as a powerful Multi Agent collaboration tool, provides developers with a convenient way to build intelligent collaboration systems...
1yrs ago
056.2K
利用视觉模型提取图片文本的OCR Prompt

OCR Prompt for Extracting Image Text Using Visual Models

When faced with complex text structures, or mixed-text content, it is good to utilize the visual model OCR capability to extract the content. Multi-modal macromodels or specialized visual models can understand the content of the image and receive instructions to perform the recognition task, we will use this ability to make the output content meet our requirements. O...
1yrs ago
056.2K
文心智能体教程:(四)加工文档并同步到知识库

Wenxin Intelligent Body Tutorial: (4) Processing Documents and Synchronizing to the Knowledge Base

Introduction to Knowledge Base The knowledge base is the data basis for the intelligent body's output answers, which is suitable for developers with professional data accumulation, and developers who have accuracy and professionalism requirements for the output results. Upload your own data in the knowledge base module, and the big model interacts with the user in the process, based on the similar content retrieved from the knowledge base...
2yrs ago
056.2K
如何让不懂编程的人低门槛使用 Windsurf 写出完整应用

How to make it low-barrier for people who don't know how to program to write complete apps using Windsurf

Don't believe the news on the Internet about "a man who doesn't know how to program and made a million dollars a month with Windsurf", he may not really know how to program, but his father did. If you don't know a little bit of programming or even English, it's really hard to write a complete program. The internet is full of people who know how to program and teach you how to use Windsurf in the real world....
1yrs ago
056.1K
杀手级流量入口:利用AI智能体为网站、公众号长期获取外部流量

Killer traffic portal: using AI intelligent body to get external traffic for websites and public numbers in the long run

At present, the domestic mainstream Internet companies are vigorously developing large models, AI intelligent body or AI assistant is an important product of the large model for consumer popularization: there are tencent yuanbao (yuan ware), xunfei starfire, wenxin intelligent body platform, wisdom spectrum light speech, buckle, etc... So AI intelligent body is an important flow...
1yrs ago
056K
OpenAI 函数调用(Function calling)

OpenAI Function calling

OpenAI Function calling V2 Features The core goal of Function calling V2 is to give OpenAI models the ability to interact with the outside world, which is mainly reflected in the following two core functions: data acquisition ...
1yrs ago
056K
Late Chunking×Milvus:如何提高RAG准确率

Late Chunking x Milvus: How to Improve RAG Accuracy

01.Background In RAG application development, the first step is to chunk the document, efficient document chunking can effectively improve the accuracy of the subsequent recall content. Efficient document chunking can effectively improve the accuracy of the subsequent recalled content. How to efficiently chunk is a hot topic of discussion, there are such as fixed-size chunking, random-size chunking...
1yrs ago
055.8K
斯坦福大学开源的ChatGPT论文写作提示词

Stanford University's open source ChatGPT essay writing prompts

In academic writing, clear, concise and persuasive expression is essential to communicate research findings. However, many non-native English-speaking researchers face language barriers when writing and embellishing academic papers. To address this problem, Stanford University has shared a series of efficient paper touch-ups through an open source project to mention...
1yrs ago
055.7K
2024年度RAG清单,RAG应用策略100+

2024 RAG Inventory, RAG Application Strategy 100+

Looking back to 2024, the big models are changing day by day, and hundreds of intelligent bodies are competing. As an important part of AI applications, RAG is also a "group of heroes and lords". At the beginning of the year, ModularRAG continues to heat up, GraphRAG shines, and in the middle of the year, open source tools are in full swing, and knowledge graphs are...
1yrs ago
055.5K
如何将copilot安装到国内电脑

How to install copilot to domestic computer

For Windows 11 users, the copilot button will not appear in the country, even if hanging ladders, for many users this is a little less convenient. However, this article can be realized through a convenient way to show the copilot on the taskbar, the use of which can be square...
1yrs ago
055.5K
Agora: a scalable multi-agent protocol for large language models

Agora: a scalable multi-agent protocol for large language models

INTRODUCTION This study introduces Agora, a metacommunication protocol designed for Large Language Model (LLM) networks.Agora aims to solve the trilemma between efficiency, scalability, and flexibility when communicating between agents by using a mixture of natural language and structured protocols to achieve efficient...
2yrs ago
055.4K
解决Suno幻觉问题

Solving the Suno hallucination

Chirp may sing slurred words at certain times.' Phantom hearing' timing is when the song deviates from the expected lyrics, or Chirp is unable to find lyrics that match the melody. We can't 'fix' the lyrics after the audio clip has been generated, but we try as much as possible to prevent the same when the audio is generated again...
2yrs ago
055.2K
让大模型在回答问题之前多思考一会的提示词,适合讨论复杂有辩证空间的问题。

Cue words that allow the large model to think a little longer before answering a question are appropriate for discussing complex issues with dialectical space.

It is always better to let the big model try to think before solving the problem, such as the COT method. However, as the ability of the big model increases, the original COT is also gradually launched into the history stage, not to mention the o1 after the emergence of the dimensionality of the blow. The communication process between human and AI allows the big model to "think"...
1yrs ago
055.2K
Kimi官方提示词:短剧脚本

Kimi's official cue word: skit scripts

You are an expert in writing scripts for popular short videos. You have a lot of creativity and IDEA, master all kinds of online popular terriers, and deeply accumulate knowledge and news information about games, fashion, clothing, fitness, food, beauty, and other popular areas on short video platforms; when short video script creation, you need to fully integrate these...
2yrs ago
055K
生成网页版PPT(演示文稿)的提示词

Tips for generating web-based PPTs (presentations)

Prompt Words # Professional Presentation Design Requirements You are a professional presentation designer and front-end development expert with a deep understanding of modern HTML presentation design trends and best practices, and are especially adept at creating RevealJS presentations with high aesthetic value. Your design work not only fe...
1yrs ago
055K
Kimi官方提示词:期刊审稿

Kimi's official cue word: journal review

I would like you to act as a journal reviewer. You will be required to review and critique submitted articles by critically assessing their research, methods, methodology, and conclusions, and by providing constructive criticism of their strengths and weaknesses. ## Note - You only start answering when a user asks a question...
2yrs ago
055K
copilot studio使用说明

copilot studio instructions

Log in to change your password Get your account number and password. Login at [https://copilotstudio.microsoft.com/] Change the initial password to your own and remember your changed password. The new password needs to contain both upper and lower case letters...
1yrs ago
054.9K
中国知名的AI产品都有哪些?

What are the well-known AI products in China?

Many well-known AI products have emerged in China in recent years, and they have demonstrated strong capabilities in various fields. According to the product function, we can roughly divide them into the following categories: First, text and content creation: Baidu - Wenxin Yiyin: an AI product that specializes in long text generation...
1yrs ago
054.9K
DeepRetrieval:强化学习驱动的高效信息检索查询生成

DeepRetrieval: efficient information retrieval query generation driven by reinforcement learning

Abstract Information retrieval systems are critical for efficient access to large document collections. Recent approaches utilize Large Language Models (LLMs) to improve retrieval performance through query augmentation, but typically rely on expensive supervised learning or distillation techniques that require significant computational resources and manually labeled data ...
1yrs ago
054.9K
Browse AI

What is the Browse AI "batch run" feature?

Our batch run feature allows you to easily perform multiple tasks at the same time. This feature is a game-changer for users who have to deal with massive amounts of data scattered across countless detail pages. Its automated operation eliminates the tedious step of manually manipulating each URL individually, saving...
2yrs ago
054.9K
OpenAI-o1有多厉害?深度优化论文,提升论文写作质量!30个极品提示词分享

How powerful is OpenAI-o1? Deeply Optimize Your Dissertation to Improve the Quality of Your Dissertation Writing! 30 Extreme Prompt Words to Share

A UCI physics PhD tested o1 and found that the code for his PhD thesis, which took him 1 year to complete, was implemented by AI in less than an hour. o1 models are already strong enough to straighten out PhD thesis code! This also means revolutionizing the writing of academic papers. By carefully constructing prompt words...
1yrs ago
054.7K
搭建RAG应用,Embedding模型如何选?

How to choose Embedding model for building RAG application?

When building a RAG system, choosing the right Embedding model is a crucial step, here are my key factors and suggestions to consider when choosing an Embedding model for your reference: Clarify the application scenarios First of all, it is necessary to clarify the specific application scenarios of the RAG system and...
1yrs ago
054.7K
扣子教程:使用”代码节点“将长文本进行分段处理

Buckle tutorial: using the "code node" will be a long text for segmentation

In the process of learning to automate novel generation, segmenting long text and then generating shot scripts for the segments is an important part of the work. This method can reduce the number of picture images generated, but it needs to be accompanied by a sentence-by-sentence generation voiceover tool. In order to ensure that the picture and sound are consistent. Many tools to first import...
2yrs ago
054.4K
将描述业务流程的文档转换为业务流程图:以辅导公司上市文档为例

Converting a document describing a business process into a business process diagram: an example of a document for coaching a company to go public

Someone in the group asked: which senior knows which ai can draw the flowchart of listed company information? Guess it is listing counseling documents related to the process, in fact, do not need any tools, as long as you can draw a sample flowchart, so that the big model to generate SVG code can be, of course, Mermaid syntax can be...
1yrs ago
054.3K
多样本(示例)越狱攻击

Multibook (example) jailbreak attack

Researchers have investigated a "jailbreak attack" technique - a method that can be used to bypass security fences set up by developers of large language models (LLMs). The technique, known as a "multisample jailbreak attack," uses Anthropic's own models, as well as those produced by other AI companies, to...
1yrs ago
054.3K
AI 教育赛道万字解析:代表性产品有哪些?机会在哪里?未来可能性?

AI education track 10,000 words analysis: what are the representative products? What are the opportunities? What are the future possibilities?

Education has long been considered one of the industries that will be changed the most by LLM. education makes up a large portion of ChatGPT's usage scenarios, and its usage often fluctuates with the start of the school year and the regularity of vacations. Andrej Karpathy has chosen education as the direction for his venture. People are expect...
1yrs ago
054.3K
用万能的“第一性原则”优化你的提示词

Optimize your prompts with the all-powerful 'firstness principle'

This method is simple but effective, the effect is comparable to the "principle of two or eight", he will not optimize the prompt word itself, but to optimize the quality of the output answer. If the cue is used to analyze the question, just insert the "firstness principle" in the appropriate place on top of the original cue. As an example, if there is no...
1yrs ago
054.3K
生成与优化React代码性能并给出合理建议的提示词

Generate and optimize React code performance and give sensible advice on prompt words

Core Functions Component-level re-rendering analysis Detect redundant rendering triggered by high-level state changes (e.g., a parent component state change triggers an unrelated child component update) Identify components that do not use React.memo and recommend optimizing child component rendering through memoization Property stability detection Discover inline objects/arrays...
1yrs ago
054.1K
将文档转换为 Mermaid 格式知识(事实关系)图谱

Convert documents to Mermaid format knowledge (factual relationship) mapping

Recently, I've been studying Ollama's API documentation, and it's hard to build up a complete relationship in my head with all the lengthy content, so I thought of generating a Mermaid-formatted relationship chart from the API documentation, which only expresses the factual relationships. Of course, you can also use it to sort out the relationships between characters, events...
1yrs ago
054.1K
chatgpt和deepseek的主要区别是什么?

What is the main difference between chatgpt and deepseek?

ChatGPT (developed by OpenAI) and DeepSeek (developed by China Depth Seeker), as the two mainstream language models, there are significant differences in technical architecture, application scenarios, language support, etc. The following is a summary of the core areas: First, the model architecture and training logic ...
1yrs ago
054.1K
如何测试使用的 DeepSeek-R1 是不是满血版?

How can I test if the DeepSeek-R1 I'm using is a full-blooded version?

DeepSeek has been continuously on fire for over a month. The official website still has a busy server, and the suspended API recharge channel is still open. During this time, the enthusiasm for local deployment and cloud deployment remains high, and various websites accessing DeepSeek have sprung up. In addition to the familiar big...
1yrs ago
054K
BM25

BM25

Introduction Why to introduce him separately, many scenarios apply GPT3 embedded vector representation, the efficiency and results may not be as good as the traditional model, which needs to be paid attention to all the time. BM25 is a vector space model, but it does not belong to the word vector model, document vector model, image vector model, knowledge graph vector...
1yrs ago
053.8K
超能画布:AI写真创作快速上手篇

Super Canvas: AI Portraits Quickstart

Step 1: Import images Note: You can click to open the folder to select, or drag and drop to upload In order to ensure the quality of the raw images and the fusion effect, it is recommended to try to choose light-colored images with simple backgrounds in Super Canvas, and to avoid images with too small a proportion of characters Step 2: Select the inspiration recipe Inspiration recipe...
2yrs ago
053.6K
为Cursor配置Critique编程提示词指令

Configuring the Critique Programming Prompt Word Directive for Cursor

Focused Analysis The purpose of this prompt is to assess and reflect on the given response, providing a rating and judging its adequacy to solve the problem or complete the task. Core elements include: Assessment Reflection: rates the adequacy, redundancy, and overall quality of the response. Scoring system: use a score from 0 to 10 to reflect on the response...
2yrs ago
053.6K
CoT(Chain-of-Thought) 链式思维

CoT (Chain-of-Thought) Chain of Thought

AKA: Chain-of-Thought Definition and Principles of Chain-of-Thought "Chain-of-Thought" (CoT for short) is a method of thinking based on the principle of creating a chain of thoughts consisting of a series of linked thinking steps. The method works by dividing the thinking process into...
2yrs ago
053.6K
20秒让你理解 DeepSeek-R1 与 ChatGPT 的差距有多大

20 seconds to understand how far DeepSeek-R1 is from ChatGPT

The most basic ability of the big model is instruction following, with the document: OpenAI o3-mini system manual (in Chinese) uploaded as an attachment to allow DeepSeek-R1 and ChatGPT to write social media blasts respectively (here I used a completely inappropriate prompt...
1yrs ago
053.6K