AI Prompt
A text input given to an AI language model to guide its output. Prompts can range from simple questions to detailed instructions with context, constraints, and formatting requirements.
Loading...
Loading...
Key terms and definitions for AI prompts, prompt engineering, and working with AI language models.
A text input given to an AI language model to guide its output. Prompts can range from simple questions to detailed instructions with context, constraints, and formatting requirements.
The practice of designing and refining prompts to get better, more accurate, or more creative outputs from AI models. Techniques include chain-of-thought, few-shot learning, and role-based prompting.
A special instruction set that defines the AI model's behavior, role, and constraints before the user conversation begins. System prompts set the tone and boundaries for all responses.
A technique where you include a few examples in your prompt to show the AI the format or style you want. This helps the model understand your expectations without explicit rules.
Giving the AI a task without any examples. The model relies entirely on its training to understand and complete the request. Simpler but less predictable than few-shot.
A prompting technique that asks the AI to show its reasoning step by step before giving a final answer. This improves accuracy on complex tasks like math, logic, and analysis.
A unit of text that AI models process. Tokens can be words, parts of words, or characters. Models have token limits that determine how much text they can process at once.
The maximum amount of text (in tokens) an AI model can process in a single request. This includes both the prompt and the response. Larger context windows allow for more complex conversations.
A setting (0-1) that controls how creative or deterministic AI output is. Lower temperatures (0-0.3) produce more focused, predictable text. Higher temperatures (0.7-1) produce more creative, varied output.
A sampling setting that controls the diversity of AI output. Top-P of 0.9 means the model considers the top 90% of probability mass when choosing the next token. Lower values make output more focused.
An AI chatbot developed by OpenAI that uses GPT (Generative Pre-trained Transformer) language models. It processes natural language prompts and generates human-like responses for tasks like writing, coding, and analysis.
An AI assistant developed by Anthropic, designed to be helpful, harmless, and honest. Claude is known for detailed analysis, long-form writing, and strong performance on complex reasoning tasks.
Google's family of AI models (formerly Bard) that can process text, images, and code. Gemini is integrated into Google products and available as a standalone chatbot.
A reusable prompt structure with placeholders for variables. Templates allow you to quickly generate customized prompts by filling in specific details like topic, audience, or tone.
A document that outlines the responsibilities, requirements, qualifications, and benefits of a job position. Well-written JDs attract qualified candidates and set clear expectations.
A one-page document submitted with a resume that introduces the applicant, highlights relevant experience, and explains why they're a good fit for the position.
A concise document listing your education, work experience, skills, and achievements. Resumes are used to apply for jobs and are typically scanned by both humans and ATS (Applicant Tracking Systems).
An AI model trained on vast amounts of text data to understand and generate human language. Examples include GPT-4, Claude, and Gemini. LLMs power chatbots, writing tools, and code assistants.
A technique that combines AI language models with external knowledge retrieval. The model first searches relevant documents, then generates responses based on that retrieved context.
The process of further training a pre-existing AI model on specific data to improve its performance on particular tasks. This customizes the model for niche use cases without building from scratch.
When an AI model generates confident-sounding but factually incorrect or fabricated information. Hallucinations happen because models predict plausible text rather than verified facts. Reducing hallucinations requires clear constraints and fact-checking.
A security vulnerability where an attacker crafts input that overrides or bypasses the original system prompt instructions. It's the AI equivalent of SQL injection — the user's input tricks the model into ignoring its original rules.
A training technique where AI models learn from human preferences by ranking outputs. Humans compare model responses, and the feedback is used to fine-tune the model toward more helpful, accurate, and safe outputs.
The practice of anchoring AI responses to verified, real-world sources rather than letting the model rely solely on its training data. Techniques include RAG, citation requirements, and providing reference documents in the prompt.
A numerical vector representation of text that captures its semantic meaning. Embeddings allow computers to compare text similarity, power search engines, and enable RAG systems to find relevant context.
A search method that uses embeddings to find semantically similar content rather than exact keyword matches. Instead of searching for the word 'happy,' a vector search can find content about 'joyful' or 'content' because they have similar meanings.
OpenAI's flagship multimodal model that can process text, images, and audio natively. GPT-4o is faster and more capable than GPT-4, with stronger performance on complex reasoning, coding, and creative tasks.
The initial instruction set provided to an AI model that defines its behavior, persona, and constraints for the entire conversation. Unlike user messages, system messages set persistent rules the model follows throughout.
An AI safety company that develops Claude, a family of large language models focused on being helpful, harmless, and honest. Anthropic pioneered Constitutional AI and is known for rigorous safety research.
The company behind ChatGPT and the GPT family of models. OpenAI develops AI research and products including GPT-4o, DALL-E, and the API platform used by millions of developers.
AI models that can process and generate multiple types of content — text, images, audio, and video — within a single interaction. Examples include GPT-4o (text + images + audio) and Gemini (text + images + code).
AI systems that can autonomously plan, execute multi-step tasks, use tools, and make decisions without constant human guidance. Unlike simple chatbots, agentic AI can browse the web, write and run code, and coordinate complex workflows.
The maximum number of tokens an AI model can process in a single request, covering both input and output. Exceeding the limit causes truncation. Common limits: GPT-4o (128K), Claude (200K), Gemini (1M+).
A parameter (0.0–2.0) that controls randomness in AI output. Lower values (0–0.3) produce focused, deterministic text. Higher values (0.7–1.0) produce more creative, varied output. Default is usually 1.0.
Software used by employers to collect, sort, and rank job applications. ATS scans resumes for keywords, formatting, and qualifications. Understanding ATS helps job seekers optimize their resumes to pass automated screening.
A brief 2–3 sentence statement at the top of a resume that summarizes your career goals and what you bring to the role. Modern resumes often replace objectives with professional summaries that focus on value to the employer.
A formal letter submitted to an employer announcing your intent to leave your position. It typically includes your last working day, a brief reason for leaving, and an offer to help with the transition.
A one-page letter submitted with a job application that introduces the candidate, highlights relevant experience, and explains why they're a strong fit for the position. Cover letters complement resumes by adding personality and context.
A technique where an AI model is given a small number of examples (typically 2–5) to learn a pattern before performing a task. In prompting, few-shot learning means including examples in your prompt to guide the model's output format.