The Long Competition is an ML competition where students solve a ML problem over the course of several months. Participate to gain ML experience and win prize money.
You will design an agent to play the social deduction game Spyfall. In this game, all players know a secret location except for the spy. The spy’s goal is to guess the location, while the other players aim to identify the spy. You’ll have access to an LLM and an embedding model, and you'll need to use your natural language processing skills to optimize your agent's strategy and win the competition.
In this reinforcement learning task, you control an agent playing "Game of Life", where your goal is to dominate the grid and defeat enemy agents.
Your task is to identify 38 objects in each scene by providing a confidence level for each image. You are given 10,000 synthetic images to train a classification model on. 10,000 additional images have been set aside for evaluation.