Next-Word Guesser
A tiny language model: it guesses the next word from the sentences it learned. Build a sentence, or teach it a new one.
Pick the first word…
How do sentences usually start?
📚 What the model learned from (20 sentences)
- I like to read books.
- I like to play football.
- I like to play with my friends.
- My friends like to play games.
- My cat likes to sleep.
- My cat likes to eat fish.
- The cat is on the mat.
- The dog is in the garden.
- The dog likes to play.
- We go to school every day.
- We go to the park after school.
- We play football in the park.
- The teacher reads a story.
- The story is about a dragon.
- The dragon likes to fly.
- I read a story about a robot.
- The robot can help people.
- Robots can learn from examples.
- A computer can follow instructions.
- I go to the library every week.
How is this like a real chatbot?
This tiny model only looks at the last word and learned from 20 sentences. A large language model looks at thousands of words before, learned from billions of sentences, and works with pieces of words called tokens. But the main idea is the same: predict what comes next. Add a new sentence above and watch the guesses change: the training data shapes what the model says.
Learn more in Inside AI →