Insights From the NeurIPS 2021 NetHack Challenge
It provides insights for the AI research community on benchmarking and comparing methods in a challenging environment, but is incremental as it reports on an existing challenge.
The report summarizes the NeurIPS 2021 NetHack Challenge, where participants developed agents to win the game NetHack, resulting in many approaches significantly beating previous best results and showing symbolic bots outperform deep RL by a large margin, though no agent came close to winning.
In this report, we summarize the takeaways from the first NeurIPS 2021 NetHack Challenge. Participants were tasked with developing a program or agent that can win (i.e., 'ascend' in) the popular dungeon-crawler game of NetHack by interacting with the NetHack Learning Environment (NLE), a scalable, procedurally generated, and challenging Gym environment for reinforcement learning (RL). The challenge showcased community-driven progress in AI with many diverse approaches significantly beating the previously best results on NetHack. Furthermore, it served as a direct comparison between neural (e.g., deep RL) and symbolic AI, as well as hybrid systems, demonstrating that on NetHack symbolic bots currently outperform deep RL by a large margin. Lastly, no agent got close to winning the game, illustrating NetHack's suitability as a long-term benchmark for AI research.