Challenge Rules

Reinforcement Learning Track: 

You may modify the model architecture, RL algorithm, and reward function. 
You may not alter the training tasks or the sampling order of the training tasks. 
You may not precompute large amounts of work, for example, through neural architecture search or massive hyperparameter sweeps that tune to multiple significant digits. 

Winners will be required to open-source their code in order to be eligible for a cash prize and co-authorship. We will retrain your submission from scratch for 8 hours on an A100 with at least 12 cores. The compute limit is intended to make this track fair for academic labs and independent researchers.
 

Curriculum Generation Track: 

You may modify the generation and sampling of tasks as well as their rewards. 
You may not alter the model architecture or RL algorithm. 
You may not precompute and upload a specific set of tasks through large scale simulation. 

Winners will be required to open-source their code in order to be eligible for a cash prize and co-authorship. We will retrain the baseline from scratch with your curriculum generator for 8 hours on an A100 with at least 12 cores. The compute limit is intended to make this track fair for academic labs and independent researchers.
 

No Holds Barred Track: 

You may modify the model architecture, RL algorithm, reward function, task generation, task sampling, etc. and are not constrained by compute. 
Upload your trained model for evaluation. 

Winners are strongly encouraged but not required to open-source their code. 
Winners will be required to disclose their general approach in order to be eligible for co-authorship.
 

General Rules for All Tracks:

1. Do not attempt to circumvent the submission limit by making multiple accounts or otherwise.
2. Do not interfere with our leaderboard by uploading submissions that violate track-specific restrictions.
3. Do not attempt to modify the stats recorded to the leaderboard or surreptitiously determine held-out tasks.
4. Do not attempt to access or modify other participants’ submissions
5. Do not write code that makes alliances with other participants' policies
6. Your participation is at our discretion. Harassing organizers, other participants, or other disruptive or rule-breaking behavior will result in a ban and forfeiture of any and all prizes.