Research map / Reinforcement learning / Visual and representation RL
The NetHack Learning Environment
Heinrich Küttler, Nantas Nardelli, Alexander Miller, Roberta Raileanu, Marco Selvatici, Edward Grefenstette, Tim Rocktäschel
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add RND exploration | IMPALA CNN · Staircase, monk | success rate: +13.58pp | Improved |
| add RND exploration | IMPALA CNN · Staircase, tourist | success rate: +6.52pp | Improved |
| add RND exploration | IMPALA CNN · Staircase, valkyrie | success rate: +16.34pp | Improved |
| add RND exploration | IMPALA CNN · Staircase, wizard | success rate: -12.96pp | Hurt |
| add RND exploration | IMPALA CNN · Pet task | success rate | Improved |
| add RND exploration | IMPALA CNN · Eat task | return | Hurt |
| add RND exploration | IMPALA CNN · Gold, monk | average dungeon level: 4.2 -> 5.0 | Improved |
| add RND exploration | IMPALA CNN · Score, monk | average score: 748 -> 780 | Improved |
| use RND intrinsic reward only, without extrinsic reward (ablation) | IMPALA CNN with RND · NetHack dungeon progress, fixed seed | dungeon progress | Hurt |
| replace the symbolic feature extractor with the original RND architecture using a random 8-dimensional symbolic embedding (ablation) | IMPALA CNN with RND · NetHack tasks | task performance | No clear effect |
| use only full-display glyph embeddings or only agent-centered crop embeddings (ablation) | IMPALA CNN with RND · NetHack tasks | task performance | No clear effect |
| train with more environment seeds | IMPALA CNN · Score and Staircase | test performance | Improved |
Closest papers by meaning
- MiniHack the Planet: A Sandbox for Open-Ended Reinforcement Learning Research (NeurIPS 2021 · similarity 0.92)
- Foundations of Multi-Agent Learning in Dynamic Environments: Where Reinforcement Learning Meets Strategic Decision-Making (AAAI 2025 · similarity 0.87)
- Enabling Adaptive Agent Training in Open-Ended Simulators by Targeting Diversity (NeurIPS 2024 · similarity 0.87)
- Prompted Policy Search: Reinforcement Learning through Linguistic and Numerical Reasoning in LLMs (NeurIPS 2025 · similarity 0.87)
- FightLadder: A Benchmark for Competitive Multi-Agent Reinforcement Learning (ICML 2024 · similarity 0.87)
- LLM-Explorer: A Plug-in Reinforcement Learning Policy Exploration Enhancement Driven by Large Language Models (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
