13 papers with code • 0 benchmarks • 0 datasets

Mean in-game score over 1000 episodes with random seeds not seen during training. See (Section 2.4 Evaluation Protocol) for details.


Use these libraries to find NetHack models and implementations

Most implemented papers

The NetHack Learning Environment

facebookresearch/nle NeurIPS 2020

Here, we present the NetHack Learning Environment (NLE), a scalable, procedurally generated, stochastic, rich, and challenging environment for RL research based on the popular single-player terminal-based roguelike game, NetHack.

BeBold: Exploration Beyond the Boundary of Explored Regions

facebookresearch/nle 15 Dec 2020

In this paper, we analyze the pros and cons of each method and propose the regulated difference of inverse visitation counts as a simple but effective criterion for IR.

CORA: Benchmarks, Baselines, and Metrics as a Platform for Continual Reinforcement Learning Agents

agi-labs/continual_rl 19 Oct 2021

In this work, we present CORA, a platform for Continual Reinforcement Learning Agents that provides benchmarks, baselines, and metrics in a single code package.

Hierarchical Kickstarting for Skill Transfer in Reinforcement Learning

ucl-dark/skillhack 23 Jul 2022

In this paper, we investigate how skills can be incorporated into the training of reinforcement learning (RL) agents in complex environments with large state-action spaces and sparse rewards.

MiniHack the Planet: A Sandbox for Open-Ended Reinforcement Learning Research

facebookresearch/minihack 27 Sep 2021

By leveraging the full set of entities and environment dynamics from NetHack, one of the richest grid-based video games, MiniHack allows designing custom RL testbeds that are fast and convenient to use.

SILG: The Multi-environment Symbolic Interactive Language Grounding Benchmark

vzhong/silg 20 Oct 2021

We hope SILG enables the community to quickly identify new methodologies for language grounding that generalize to a diverse set of environments and their associated challenges.

NovelD: A Simple yet Effective Exploration Criterion

tianjunz/NovelD NeurIPS 2021

We analyze NovelD thoroughly in MiniGrid and found that empirically it helps the agent explore the environment more uniformly with a focus on exploring beyond the boundary.

Insights From the NeurIPS 2021 NetHack Challenge

dllllb/neurips2021-nethack-raph 22 Mar 2022

In this report, we summarize the takeaways from the first NeurIPS 2021 NetHack Challenge.

Improving Policy Learning via Language Dynamics Distillation

vzhong/language-dynamics-distillation 30 Sep 2022

Recent work has shown that augmenting environments with language descriptions improves policy learning.

Dungeons and Data: A Large-Scale NetHack Dataset

facebookresearch/nle 1 Nov 2022

Recent breakthroughs in the development of agents to solve challenging sequential decision making problems such as Go, StarCraft, or DOTA, have relied on both simulated environments and large-scale datasets.