Voyager: An Open-Ended Embodied Agent with Large Language Models
Guanzhi Wang, Yuqi Xie, Yunfan Jiang, Ajay Mandlekar, Chaowei Xiao,, Yuke Zhu, Linxi Fan, Anima Anandkumar

TL;DR
Voyager is an LLM-powered embodied agent in Minecraft that continuously explores, learns, and adapts through an innovative curriculum, skill library, and iterative prompting, achieving superior performance and generalization without fine-tuning.
Contribution
It introduces Voyager, a novel open-ended embodied agent leveraging large language models for lifelong learning and exploration in Minecraft, with a new iterative prompting mechanism and skill library.
Findings
Voyager achieves 3.3x more unique items than prior methods.
It travels 2.3x longer distances in the environment.
It unlocks key milestones up to 15.3x faster than state-of-the-art.
Abstract
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) an automatic curriculum that maximizes exploration, 2) an ever-growing skill library of executable code for storing and retrieving complex behaviors, and 3) a new iterative prompting mechanism that incorporates environment feedback, execution errors, and self-verification for program improvement. Voyager interacts with GPT-4 via blackbox queries, which bypasses the need for model parameter fine-tuning. The skills developed by Voyager are temporally extended, interpretable, and compositional, which compounds the agent's abilities rapidly and alleviates catastrophic forgetting. Empirically, Voyager shows strong in-context lifelong learning…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMultimodal Machine Learning Applications · Topic Modeling · Artificial Intelligence in Games
MethodsLib · Multi-Head Attention · Attention Is All You Need · Softmax · Layer Normalization · Byte Pair Encoding · Dropout · Linear Layer · Label Smoothing · Adam
