Simulations

Multi Armed Bandits

Dense Reward Random Single-Agent

Repeatedly select from a series of One-Armed Bandits, to achieve the best return over time.

The Coin Game

Competitive Deterministic Multi-Agent Perfect Information Sparse Reward

A range of two player puzzles where the aim is to force your opponent to take the last coin.

High / Low Cards

Dense Reward Partial Observation Random Single-Agent

Choose weather the next card will be ‘Higher’ or ‘Lower’ than the last.

Maze

Dense Reward Partial Observation Random Single-Agent

Explore a maze to find the exit then exploit the biases in the generator.

Mine Hunter

Partial Observation Random Single-Agent Sparse Reward

Use logic to defuse a minefield.

Twisty Puzzles

Perfect Information Random Single-Agent Sparse Reward

Traditional logic puzzles of various sizes.