AI & Machine Learning Terms Explained
A plain-language reference for the most important concepts in artificial intelligence, machine learning, and generative AI. From agents to zero-shot learning.
Agent
An AI system that can autonomously perceive its environment, make decisions, and take actions to achieve specified goals. AI agents often combine language models with tool-use capabilities to complete complex, multi-step tasks without continuous human guidance.
Find related toolsAgentic AI
A paradigm where AI systems act as goal-driven agents that can plan, use tools, browse the web, write and execute code, and orchestrate multi-step workflows. Agentic AI represents a shift from single-turn Q&A toward persistent, autonomous task completion.
Find related toolsArtificial Intelligence
The broad field of computer science focused on creating systems capable of performing tasks that typically require human intelligence, such as understanding language, recognizing images, making decisions, and solving problems.
Find related toolsAttention Mechanism
A component of neural networks that allows the model to focus on the most relevant parts of the input when producing output. Self-attention in transformers lets every token attend to every other token, enabling the model to capture long-range dependencies in text.
Find related toolsAutonomous AI
AI systems designed to operate independently with minimal human intervention. These systems can set sub-goals, plan actions, execute tasks, and self-correct, making them suitable for complex workflows like software development, research, and business automation.
Find related toolsBenchmark
A standardized test or dataset used to evaluate and compare AI model performance. Common benchmarks include MMLU (general knowledge), HumanEval (coding), GSM8K (math), and HellaSwag (common sense). Benchmarks help track progress but may not fully reflect real-world performance.
Find related toolsBias
Systematic errors or prejudices in AI model outputs caused by imbalanced or unrepresentative training data, flawed labeling, or problematic model design. Bias can lead to unfair or discriminatory results across different demographic groups.
Find related toolsChain of Thought
A prompting technique that encourages AI models to break down complex problems into intermediate reasoning steps before arriving at a final answer. This approach significantly improves accuracy on math, logic, and multi-step reasoning tasks.
Find related toolsCLIP (Contrastive Language-Image Pre-training)
A model developed by OpenAI that learns to connect images and text by training on image-text pairs from the internet. CLIP enables zero-shot image classification and is a key building block in text-to-image systems like DALL-E and Stable Diffusion.
Find related toolsComputer Vision
A field of AI that enables machines to interpret and understand visual information from images and videos. Applications include object detection, facial recognition, medical imaging analysis, autonomous driving, and visual quality inspection.
Find related toolsContext Window
The maximum amount of text (measured in tokens) that a language model can process in a single interaction. Larger context windows allow models to handle longer documents and maintain coherence over extended conversations. Modern models range from 4K to over 1M tokens.
Find related toolsData Augmentation
Techniques for artificially expanding a training dataset by creating modified versions of existing data. In image AI, this includes rotations, flips, and color changes; in text AI, it includes paraphrasing, back-translation, and synonym substitution.
Find related toolsDeep Learning
A subset of machine learning that uses artificial neural networks with multiple layers (hence 'deep') to learn complex patterns from data. Deep learning powers most modern AI breakthroughs, including image recognition, language understanding, and generative AI.
Find related toolsDiffusion Model
A type of generative AI model that creates images by gradually removing noise from a random starting point. Models like Stable Diffusion, DALL-E, and Midjourney use this approach to generate high-quality images from text descriptions.
Find related toolsEdge AI
AI processing that occurs directly on local devices (smartphones, IoT sensors, cameras) rather than in the cloud. Edge AI offers faster response times, better privacy, and reduced bandwidth usage, making it ideal for real-time applications.
Find related toolsEmbedding
A numerical representation of data (text, images, or audio) as a dense vector of numbers. Embeddings capture semantic meaning, allowing AI systems to measure similarity between concepts. They are foundational for search engines, recommendation systems, and RAG pipelines.
Find related toolsEmergent Behavior
Unexpected capabilities that arise in large AI models as they scale up, which were not explicitly programmed or anticipated. Examples include in-context learning, chain-of-thought reasoning, and the ability to perform tasks not seen during training.
Find related toolsFederated Learning
A machine learning approach where models are trained across multiple decentralized devices or servers without sharing raw data. This preserves data privacy while still enabling collaborative model improvement, commonly used in healthcare and mobile applications.
Find related toolsFew-Shot Learning
The ability of an AI model to learn a new task from only a handful of examples, typically provided within the prompt. This contrasts with traditional machine learning, which requires thousands or millions of training examples.
Find related toolsFine-Tuning
The process of further training a pre-trained AI model on a specific, smaller dataset to adapt it for a particular task or domain. Fine-tuning allows organizations to customize general-purpose models for specialized applications like legal analysis, medical coding, or brand-specific content.
Find related toolsFoundation Model
A large AI model trained on broad, diverse data that can be adapted to a wide range of downstream tasks. Examples include GPT-4, Claude, Gemini, and Llama. Foundation models serve as the base layer upon which specialized applications are built.
Find related toolsGAN (Generative Adversarial Network)
A type of generative AI architecture consisting of two neural networks — a generator that creates content and a discriminator that evaluates it — competing against each other to produce increasingly realistic outputs. GANs were pioneering in AI image generation before diffusion models became dominant.
Find related toolsGenerative AI
AI systems that can create new content — including text, images, music, code, and video — based on patterns learned from training data. Generative AI encompasses large language models, diffusion models, and other architectures that produce novel outputs rather than just classifying or predicting.
Find related toolsGPT (Generative Pre-trained Transformer)
A family of large language models developed by OpenAI, built on the transformer architecture. GPT models are pre-trained on vast text corpora to predict the next token, enabling them to generate human-like text, answer questions, write code, and perform diverse language tasks.
Find related toolsGrounding
The practice of connecting AI model outputs to verified, factual information sources to reduce hallucinations and improve accuracy. Grounding techniques include RAG, web search integration, and citation of source documents within generated responses.
Find related toolsHallucination
When an AI model generates plausible-sounding but factually incorrect or fabricated information. Hallucinations are a key challenge with large language models and can include made-up citations, fictional events, or incorrect technical details presented with high confidence.
Find related toolsInference
The process of using a trained AI model to generate predictions or outputs from new input data. Inference is the production phase of AI, as opposed to training. Optimizing inference speed and cost is critical for deploying AI applications at scale.
Find related toolsLarge Language Model (LLM)
An AI model trained on massive amounts of text data, typically with billions or trillions of parameters, capable of understanding and generating human language. LLMs like GPT-4, Claude, Gemini, and Llama power chatbots, coding assistants, writing tools, and many other AI applications.
Find related toolsLoRA (Low-Rank Adaptation)
An efficient fine-tuning technique that adds small, trainable adapter layers to a frozen pre-trained model instead of updating all parameters. LoRA dramatically reduces the computational resources and data needed for customization, making fine-tuning accessible on consumer hardware.
Find related toolsMachine Learning
A subset of artificial intelligence where systems learn patterns from data and improve their performance over time without being explicitly programmed. Machine learning encompasses supervised learning, unsupervised learning, and reinforcement learning approaches.
Find related toolsMixture of Experts (MoE)
A model architecture where multiple specialized sub-networks (experts) exist within a single model, and a gating mechanism routes each input to the most relevant experts. MoE enables much larger models with lower inference costs, as only a fraction of parameters activate per query.
Find related toolsMultimodal AI
AI systems that can process, understand, and generate multiple types of data — such as text, images, audio, and video — within a single model. Multimodal models like GPT-4o and Gemini can analyze images, transcribe speech, and produce mixed-media responses.
Find related toolsNatural Language Processing (NLP)
A branch of AI focused on enabling computers to understand, interpret, and generate human language. NLP powers applications like chatbots, translation services, sentiment analysis, text summarization, and voice assistants.
Find related toolsNeural Network
A computing system inspired by the structure of the human brain, consisting of interconnected nodes (neurons) organized in layers. Neural networks process data by passing signals through these layers, learning to recognize patterns and make decisions through training.
Find related toolsOpen Source AI
AI models and tools released with publicly available weights, code, and documentation, allowing anyone to use, modify, and distribute them. Notable open-source models include Llama, Mistral, and Stable Diffusion. Open-source AI democratizes access and enables customization.
Find related toolsParameter
A learnable value within a neural network that gets adjusted during training. The number of parameters (often in the billions for LLMs) is a rough indicator of a model's capacity. GPT-4 is estimated to have over a trillion parameters, while smaller models may have 7-70 billion.
Find related toolsPrompt Engineering
The practice of crafting effective instructions and context to elicit desired outputs from AI models. Techniques include providing examples (few-shot), specifying output formats, assigning personas, using chain-of-thought reasoning, and structuring multi-step instructions.
Find related toolsQuantization
A technique for reducing the precision of a model's numerical values (e.g., from 32-bit to 4-bit) to decrease memory usage and speed up inference. Quantization makes it possible to run large AI models on consumer GPUs and edge devices with minimal quality loss.
Find related toolsRAG (Retrieval-Augmented Generation)
An architecture that enhances AI model responses by first retrieving relevant information from external knowledge sources (databases, documents, APIs) and then using that context to generate more accurate and grounded answers. RAG reduces hallucinations and keeps outputs up to date.
Find related toolsReinforcement Learning
A machine learning paradigm where an agent learns to make decisions by taking actions in an environment and receiving rewards or penalties. It's used in game-playing AI, robotics, recommendation systems, and for aligning language models with human preferences (RLHF).
Find related toolsRLHF (Reinforcement Learning from Human Feedback)
A training technique where human evaluators rank AI model outputs, and the rankings are used to train a reward model that guides further optimization. RLHF is a key method for aligning language models with human values, improving helpfulness, and reducing harmful outputs.
Find related toolsSemantic Search
A search approach that understands the meaning and intent behind queries rather than just matching keywords. Powered by embeddings and vector databases, semantic search can find conceptually relevant results even when exact terms don't match.
Find related toolsSynthetic Data
Artificially generated data used to train or evaluate AI models. Synthetic data can supplement real-world data when it is scarce, expensive, or privacy-sensitive. AI models themselves are increasingly used to generate high-quality synthetic training data for other models.
Find related toolsTemperature
A parameter that controls the randomness of an AI model's output. Lower temperatures (e.g., 0.1) produce more deterministic and focused responses, while higher temperatures (e.g., 1.0) produce more creative and varied outputs. Temperature is a key setting for tuning AI behavior.
Find related toolsToken
The basic unit of text that language models process. A token can be a word, part of a word, or a punctuation mark. For English text, one token is roughly 3-4 characters. Token counts determine input limits, output lengths, and API pricing for LLM services.
Find related toolsTop-p Sampling (Nucleus Sampling)
A text generation strategy where the model considers only the smallest set of next-token candidates whose cumulative probability exceeds a threshold p. Top-p sampling balances diversity and quality, often used alongside or instead of temperature for controlling output randomness.
Find related toolsTraining Data
The dataset used to teach an AI model patterns and knowledge. The quality, size, diversity, and representativeness of training data directly impact model performance, biases, and capabilities. Curating high-quality training data is one of the most important steps in AI development.
Find related toolsTransfer Learning
A technique where a model trained on one task is repurposed as the starting point for a different but related task. Transfer learning allows developers to leverage pre-trained models and adapt them with much less data and compute than training from scratch.
Find related toolsTransformer
A neural network architecture introduced in 2017 that uses self-attention mechanisms to process input data in parallel rather than sequentially. Transformers are the foundation of virtually all modern large language models and have revolutionized NLP, computer vision, and generative AI.
Find related toolsVector Database
A specialized database designed to store, index, and query high-dimensional vector embeddings efficiently. Vector databases like Pinecone, Weaviate, and ChromaDB are essential infrastructure for semantic search, recommendation systems, and RAG applications.
Find related toolsZero-Shot Learning
The ability of an AI model to perform a task it was never explicitly trained on, without any task-specific examples. Large language models exhibit strong zero-shot capabilities, allowing them to follow new instructions, classify text, or translate languages without prior fine-tuning.
Find related toolsExplore AI Tools by Category
Now that you know the terminology, discover 1,300+ AI tools across 20+ categories — from writing assistants to code generators.