#Cosmic Carnage: A Reinforcement Learning Challenge 

Cosmic Carnage is a progressive learning challenge consisting of 16 independent levels. Participants must design an intelligent agent that learns to solve each level through training.  While each level defines a separate environment and is independent, participants are allowed to reuse or share model weights across levels if they choose to do so.

##Competition Rules & Guidelines:

- The competition runs for a duration of three days, during which you must complete as many levels as possible.

- There is no restriction on the type of learning algorithm used. Any approach is allowed as long as the solution is learning-based.

- Hardcoded solutions or direct analytical solvers are strictly prohibited.

- The agent must learn through training and produce a visible learning curve.

- Use of artificial intelligence tools, including large language models, is strongly encouraged.