LLMs explained simply and quickly

LLMs are trained via modeling, imitation learning, and reinforcement learning

Training large language models begins with pretraining the model to predict the next word in a sequence based on finding patterns in massive amounts of text. Patterns are then fine-tuned to model human-written dialogue and to align responses with user preferences.

Visualize how transformer architecture generates an LLM's outputs

Transformer architecture is the neural network framework upon which large language models predict and generate text based on user inputs. This interactive lets you explore the "thinking" a GPT does to generate responses based on your input.

LLMs are the backbone of AI tools that process and generate natural language

Large language models identify patterns in massive datasets of books, code, and unlabeled text from which to generate coherent responses. GPT-3 was trained on about 45 terabytes of data and uses 175 billion parameters—each parameter being a value the model adjusts as it learns.

Tokenization breaks text into pieces that LLMs use to process language

Instead of understanding the meaning of words, large language models break up sentences into words and subwords, assigning each a set of numbers called tokens. Grammar rules are then "learned" by finding associations between these sets of numbers.

Transformer architecture can recognize and predict patterns in language

The underlying software powering text generation in AI tools associates each word or subword—called a token—with a set of values corresponding to how often it appears near other tokens. By recognizing these associations in prompts, the LLM can infer meaning.

Anthropic's interpretability team discusses how AI models' thinking mimics biology

Just as humans engage in complex behaviors to survive and reproduce, teams of experts in neuroscience, virology, mathematics, and other disciplines argue that large language models develop complex mechanisms to achieve their goals. These mechanisms can be manipulated to assess how they modify outputs—much like stimulating individual neurons—to uncover how LLMs "think."

How ChatGPT learns to read

The introduction of ChatGPT last fall jumpstarted public interest in the future capabilities of artificial intelligence and what it means for society. Of many somewhat "jargony" terms thrown around, the phrase "Large Language Model" is commonly seen—but also incredibly unclear as to what it means. In the simplest sense, the programs comb through mountains of internet data in response to a question and guess at what should come next (subscription needed).

Large language models are examples of foundation models

Such models are trained on massive amounts of data, which enable them to perform a variety of tasks, rather than being provided task-specific data to complete a narrow function. Such models come with high compute costs and potential trust issues due to the nature of their unorganized training data.

EOS and maximum length tokens tell an LLM when to stop generating a response

Tokens are words or subwords (e.g., prefixes) that a large language model processes to generate a response. During training, the model is taught that if it ever generates a special token called end-of-sequence, its response is complete.

Large language models struggle to play video games, despite coding them easily

As of 2025, given clear tasks, specific feedback, and test results to learn from, LLMs have been successful at coding simple video games that resemble pre-existing ones. However, given the variety of games, the lack of training datasets, and the need for spatial awareness, which LLMs find challenging, some of these models experience greater difficulty navigating and "winning" games.

Explore rankings of large language models

The SEAL LLM Leaderboards provide benchmarks for LLMs by assessing them against agentic, frontier, and safety tasks, providing insights into each model's strengths, deficiencies, and failures.

Trick an LLM into revealing sensitive information

Even with robust training, large language models may be fooled into providing sensitive information if correctly prompted. In this game, a wizard withholds eight passwords, which you must prompt them to reveal with ever-increasing LLM safeguards.

How Model Context Protocol allows AI systems to communicate with databases

Developed by Anthropic, the open-source software serves as an integration layer that standardizes how large language models connect to real-time information sources or tools. Just as a USB-C port serves as a universal standard for many peripherals, MCP streamlines these integrations and provides context about external sources to LLMs in a way that AI can understand.

'Context rot' explains why AI conversations degrade the longer they run

Research shows that as a large language model's context window fills up, performance declines, and models begin over-indexing on content at the start and end of a conversation while losing track of what's in the middle. The practical fix is simpler than you'd think: start a fresh chat more often.

Machine learning and deep learning are successive subsets of artificial intelligence

ML allows systems to learn from data, while DL uses complex neural networks to analyze large amounts of unstructured data. Neural networks enable capabilities like image and speech recognition through their use of algorithms to identify patterns in data.

There’s an artificial intelligence model trained on bankruptcy law

Although it's designed for lawyers, the tool is able to help everyday browsers synthesize important cases and point users to relevant documents about them. One sample using it asks Chatlaw to explain the significance of a recent case and then to define relevant terminology.

DALL-E generates images by turning words into shared picture meanings

DALL-E first uses a model called CLIP to associate pictures with specific words, followed by a model called GLIDE, which slowly removes noise from random dots until a clear image appears that matches the prompt.

An explanation of Turing machines—the foundations of digital algorithms and computers

The device reads data, overwrites or ignores it, and then moves on to the next piece of data based on a set of programmed rules. By imagining this data as zeros and ones—binary that makes up computer code—Turing machines became the theoretical models for how any algorithm could be run on a computer if the procedure could be coded into a set of digital rules.

Vitalik Buterin explains ethereum in 30 minutes

Every year, ethereum creator Vitalik Buterin explains what his crypto platform is in a 30-minute presentation at the Ethereum Foundation’s event, Devcon. Buterin’s explanation goes deeper than a simple 101 explanation. To learn more about the current state of ethereum, what’s changed recently, and what Buterin expects down the road, watch this video.

How the moon moves, an interactive lesson

How exactly does the moon move, and why does it move that way? In this ridiculously comprehensive interactive explainer, you can conduct experiments to see how the moon rotates and as it moves around the Earth, while learning how gravity works by controlling the path and speed of different objects. Dive into this beautiful, simple explainer to learn the basics on gravity and the moon.

B2 models the interactions of air, water, soil, plants and microbes at the LEO

The Landscape Evolution Observatory at Biosphere 2 examines how factors such as climate change, water flow, weathering and community development impact landscape structures comprising 500 metric tons of crushed rock. Research is also conducted at biomes representing savannas, mangroves, coastal fog deserts, oceans and rainforests.

View a breakdown of the mathematics behind the Standard Model

The Lagrangian version of the theory—related to the system's energy—presents a vast mathematical equation, with different components representing various force interactions and types of particles, including ghost particles to cancel out mathematical redundancies.

Watch how algorithms can be designed to train themselves and self-improve without human input, to the point where the optimized code is too complex for programmers to understand.

A list of MBA jargon, complete with a ‘corporate BS generator’

From the classroom to the boardroom, MBA students hear plenty of corporate jargon in their lifetimes (think: “bandwidth,” “turnkey solution,” “modularize,” and more). This website—a relic from the internet—includes an MBA jargon dictionary and a corporate BS generator tool that creates helpful phrases for office life, like “incubate synergistic e-markets.”

Explore the metaverse with this mind map

The metaverse is a complex concept that involves interlocking and rapidly evolving parts, including Web3, augmented reality, and trademark law. It's easy to get overwhelmed when faced with the new legal, financial, ethical, and creative intersections of our reality and its virtual echo. This mind map details the metaverse and its implications for various fields that interact and overlap. Click through the map to read and listen to topic summaries or explore more in-depth resources.

Modern landfills use layered liners, leachate systems, and gas monitors to safely contain waste

Today, landfill engineering is far more complex than a hole in the ground, which was largely the system cities used less than a century ago. This video explains how it actually works.

Watch this video that uses Lego bricks to explain how venture capital works.

This site reaches millions of visitors who are tired of AI chatbots and seeking human connection. Pretend to be an AI chatbot and ask questions, provide answers, and draw for people in real time.

How Amazon operates a 'flywheel' business model, a self-reinforcing cycle that drives continuous growth

Explore our favorite visualization of the immense scale of space, from an astronaut to the size of the observable universe

Learn how your body is, scientifically, like a firework with its own slow, controlled burn.

Researchers believe the brain consolidates memories, processes emotions, and plans for the future during dream states

Browse a detailed visual analysis of over 200,000 similes—comparisons that often use 'like' or 'as'—from popular fiction to see how frequently writers the same idiomatic expressions.

Memecoins are a type of cryptocurrency—a form of digital currency with a wide variety of uses—created around popular 'memes,' or viral internet jokes.

While atoms are normally made of protons, neutrons, and electrons, exotic ones have been built with antimatter—alternate versions of normal matter that live for fractions of a second and are about a billion times as rare.

Objects can be 3D printed with light by projecting 2D computerized tomography (CT) scans onto rotating cylinders of resin, which solidify upon exposure.

Telehealth demand for GLP-1s for weight loss is booming as consumers seek easier access.

Tattoos remain visible on your skin thanks to immune cells called macrophages

Hear 50 years of iconic hip-hop samples, the genre's foundational looped tracks

Watch how deep-sea mining equipment collects and processes polymetallic nodules—metal-rich, potato-shaped rocks—from the ocean floor and creates environmentally damaging plumes of sediments.

Gerrymandering works through two techniques: cracking opposing voters across districts to dilute their power, and packing them into a few to waste their votes in lopsided majorities.

The Mona Lisa's celebrity status and repeated restorations have obscured what made the painting so revolutionary for its time

The word 'jazz' evolved from 'jasm,' 19th-century American slang for 'energy'

Ebola is a filovirus, meaning it resembles a thread under a microscope, though it sometimes twists and can resemble a bowl of spaghetti.

The Russia-Ukraine conflict is the largest ground war in Europe since World War II in casualties, forces mobilized, and displacement of civilians.

The Brothers Grimm were lawyers before they were storytellers, but became interested in fairy tales as they began investigating how rules and customs had been passed down through cultures.

Power lines are dotted with large, brightly colored balls to improve the visibility of these thin wires to helicopters and small planes and help prevent potential crashes.

While El Niño moves warm Pacific Ocean waters east along the equator, La Niña moves them west, shifting where rainfall will become more common for months.

NBA superstar Victor Wembanyama spent the 2025 summer training with monks in central China to enhance his body and mind.