Learning to Modulate pre-trained Models in RL | Read Paper on Bytez