SOTAVerified

NetHack

Mean in-game score over 1000 episodes with random seeds not seen during training. See https://arxiv.org/abs/2006.13760 (Section 2.4 Evaluation Protocol) for details.

Papers

Showing 110 of 28 papers

TitleStatusHype
MaestroMotif: Skill Design from Artificial Intelligence Feedback0
BALROG: Benchmarking Agentic LLM and VLM Reasoning On Games0
Syllabus: Portable Curricula for Reinforcement Learning AgentsCode2
Online Intrinsic Rewards for Decision Making Agents from Large Language Model FeedbackCode1
PufferLib: Making Reinforcement Learning Libraries and Environments Play NiceCode4
Playing NetHack with LLMs: Potential & Limitations as Zero-Shot AgentsCode1
Craftax: A Lightning-Fast Benchmark for Open-Ended Reinforcement LearningCode3
Skill Set Optimization: Reinforcing Language Model Behavior via Transferable SkillsCode1
Fine-tuning Reinforcement Learning Models is Secretly a Forgetting Mitigation ProblemCode0
diff History for Neural Language AgentsCode1
Show:102550
← PrevPage 1 of 3Next →

No leaderboard results yet.