Thesis version

Freezing the code as of the (probably) last PhD paper. After this, I might start decoupling the RL code from the thesis-specific elements to make it more widely usable.

Reward exploration

For tracking purposes

CrowdAI-v6

This version of CrowdAI was used in the "Understanding Reinforcement Learned Crowd" paper presented at MIG 2022, published in Computers & Graphics (due for publication as of September 2022)

CrowdAI-v5

This version was used for MCMES