Click a word. The glow is how strongly this toy model looks at each other word — a sketch of the mechanism transformers actually use.
Click a word. Highlight intensity is how much this toy model "looks at" each other word while reading the clicked one. Real models use many attention heads; this is one hand-tuned row so the pronoun trick is obvious.
While reading it, strongest looks
English speakers usually take "it" as the trophy (too large to fit). The suitcase is the other candidate. Attention is how transformers keep both options in play, then pick.
Not a search engine
Attention is how each token in a sentence gathers context from the others before predicting what comes next.
Pronouns need a target
“It” is empty until the model decides whether it points at the trophy or the suitcase. That choice is attention plus the rest of the network.
Many heads, not one
Real models run this lookup in parallel many times (heads) and many layers. One colorful row is enough to feel the idea.
Transformer architecture, explained
Attention in the stack that actually ships.
Sparse attention and long context
What changes when you cannot afford full pairwise looks.
Try the Neural Network Playground →
The weighted connections under a simpler diagram.
Try Generative AI →
After attention, the model still has to pick the next word.
No. Weights were written for this sentence so the trophy/suitcase split is visible. Production models learn thousands of these patterns from data.
The architecture behind GPT, Claude, and Gemini. Attention is the part that lets each word look at other words instead of reading left-to-right only.
No. The sentence and weights are bundled in the page. Clicks never leave your browser.
Tokenizer Playground
Type anything and watch it split into tokens live — the units LLMs actually read and get billed for.
Context Window Visualizer
See how much of a model's context window your text fills up, and what happens when you run out of room.
FROG in a Bowl Prompt Builder
Fill in Format, Role, Objective, Goal, and Context — get a copy-ready structured prompt in seconds.
Machine Learning Types
Three tiny games show what supervised, unsupervised, and reinforcement learning actually mean.
Neural Network Playground
Drag two sliders and watch a real, tiny neural network turn them into a decision, live.
Generative AI Playground
Watch AI write one word at a time by predicting what's most likely to come next.
RAG Playground
Ask a question, watch notes get retrieved, then see a grounded answer versus a hallucination.
Embedding Map
Click two words and see why similar meanings sit near each other — the idea behind vector search.
Prompt Injection Lab
Watch a pasted email try to override a support agent, then flip a switch that treats it as data.
