AI Wiki

A neural network trained on massive text corpora to predict the next token, used for chat, coding, reasoning and as the brain inside AI agents.

An LLM-based system that can plan, use tools and take multi-step actions toward a goal — not just answer a single prompt.

AI systems that act autonomously over many steps to achieve a goal, combining planning, tool use and memory.

A technique that retrieves relevant documents at query time and feeds them into the LLM's prompt, so the model can answer from your data instead of memorising it.

Continuing to train a pre-trained model on your own data to specialise its behaviour, tone or domain knowledge.

The maximum number of tokens an LLM can read in a single request, including the prompt, retrieved documents and the model's own reply.

The atomic units that LLMs read and write — sub-word pieces produced by a tokenizer. Pricing and context limits are measured in tokens, not words.

The practice of designing inputs to LLMs to reliably produce useful outputs — through structure, examples, role-setting and constraints.

AI models that natively accept and/or produce more than one modality — text, image, audio, video — in a single model.

The neural-network architecture (Vaswani et al., 2017) that powers virtually every modern LLM, based on self-attention instead of recurrence.

An open standard from Anthropic for connecting LLMs to external tools and data sources through a uniform server interface.

The first algorithmically trained neural network, invented by Frank Rosenblatt in 1957, and the foundational unit that sparked the deep-learning revolution.

The conceptual bridge between Rosenblatt's hard binary perceptron and modern, differentiable, probabilistic neural networks — the soft engine that powers today's AI.