Papers with Code 아카이브
NetHack
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
- 하위 태스크
- 1
- 평가표
- 0
- 보존된 평가 기록
- 0
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.
논문
이 태스크와 하위 분야의 관련 논문입니다. 실제 성능 비교는 평가 결과에서 확인하세요.
BALROG: Benchmarking Agentic LLM and VLM Reasoning On Games
Syllabus: Portable Curricula for Reinforcement Learning Agents
Online Intrinsic Rewards for Decision Making Agents from Large Language Model Feedback
PufferLib: Making Reinforcement Learning Libraries and Environments Play Nice
Playing NetHack with LLMs: Potential & Limitations as Zero-Shot Agents
Craftax: A Lightning-Fast Benchmark for Open-Ended Reinforcement Learning
Fine-tuning Reinforcement Learning Models is Secretly a Forgetting Mitigation Problem
Skill Set Optimization: Reinforcing Language Model Behavior via Transferable Skills
diff History for Neural Language Agents
Motif: Intrinsic Motivation from Artificial Intelligence Feedback
Selective Perception: Optimizing State Descriptions with Reinforcement Learning for Language Model Actors
Scaling Laws for Imitation Learning in Single-Agent Games
LuckyMera: a Modular AI Framework for Building Hybrid NetHack Agents
Katakomba: Tools and Benchmarks for Data-Driven NetHack
Accelerating exploration and representation learning with offline pre-training
Dungeons and Data: A Large-Scale NetHack Dataset
Improving Policy Learning via Language Dynamics Distillation
Hierarchical Kickstarting for Skill Transfer in Reinforcement Learning
Insights From the NeurIPS 2021 NetHack Challenge
SILG: The Multi-environment Symbolic Interactive Language Grounding Benchmark
CORA: Benchmarks, Baselines, and Metrics as a Platform for Continual Reinforcement Learning Agents
MiniHack the Planet: A Sandbox for Open-Ended Reinforcement Learning Research
BeBold: Exploration Beyond the Boundary of Explored Regions
The NetHack Learning Environment
Exploration in NetHack With Secret Discovery
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.