Open-world survival games pose significant challenges for AI algorithms due to their multi-tasking, deep exploration, and goal prioritization requirements. Despite reinforcement learning (RL) being popular for solving games, its high sample complexity limits its effectiveness in complex open-world games like Crafter or Minecraft. We propose a novel approach, SPRING, to read Crafter's original academic paper and use the knowledge learned to reason and play the game through a large language model (LLM).

The capability of generalizing to out-of-distribution data is crucial for the deployment of machine learning models in the real world.

Specifically, we explore VPG transfer across different LLM sizes ( e .