2022/03/22 by Eric Hambro, Sharada P. Mohanty, Sharada Mohanty +57 · 1 voice
Computer Science · Mathematics · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Neural and Evolutionary Computing (cs.NE) #Reinforcement Learning in Robotics #Symbolic Computation (cs.SC) #cs.AI #cs.LG #cs.NE #cs.SC #stat.ML
paper · pdf · doi:10.48550/arxiv.2203.11889
openalex publication_date 2022/03/22 · arxiv published 2022/03/22 · arxiv updated 2022/03/22 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
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.