Technology · Agentic → Frameworks · Open Source · wiki:deep
Voyager is an LLM-powered embodied lifelong learning agent (Minecraft) with automatic curriculum, skill library of executable code, and iterative prompting. For this OS research, the valuable shape is SkillManager / skill library — not Minecraft itself. An Open-Ended Embodied Agent with Large Language Models Website(https://voyager.minedojo.org/) Arxiv(https://arxiv.org/abs/2305.16291) PDF(https://voyager.minedojo.org/assets/documents/voyager.pdf) Tweet(https://twitter.com/DrJimFan/status/1662115266933972993?s=20) https://github.com/MineDojo/Voyager/assets/25460983/ce29f45b-43a5-4399-8fd8-5dd105fd64f2 We introduce Voyager, the first LLM-powered embodied lifelong learning agent in Minecraft that continuously explores the world, acquires diverse skills, and makes novel discoveries without human intervention. Voyager consists of three key components: 1).
In this catalog it is the Frameworks option when you need to compose multi-step agent loops and tool handoffs. Keep Voyager on the observe/propose side of the loop: useful signals and drafts, with human or policy authority still required before anything irreversible runs.
Voyager’s research shape is an embodied lifelong learner: automatic curriculum, a skill library of executable code, and iterative prompting (Minecraft).
Related: smolagents · topics/33-skill-registry · topics/12-central-nervous-system
Claims below are backed by science sources on disk.
Primary / product role
MineDojo/Voyager · MODERATE
“An Open-Ended Embodied Agent with Large Language Models”
Docs or contrast
OpenHands contrast · MODERATE
“The self-hosted developer control center for coding agents and automations.”
Additional source
Voyager: An Open-Ended Embodied Agent with Large Language Models · MODERATE
“Voyager interacts with GPT-4 via blackbox queries, which bypasses the need for model parameter fine-tuning.”
9 tags · 39 out · 40 in · 3 artifacts · 0 gaps · 25 corpus docs
map_edge · 14
tech_features · 1
tech_quote · 9
tech_readme · 1