{"work":{"id":"ffe0d207-86cf-4742-a100-e988ac8b9676","openalex_id":null,"doi":null,"arxiv_id":"2305.16291","raw_key":null,"title":"Voyager: An Open-Ended Embodied Agent with Large Language Models","authors":null,"authors_text":"Guanzhi Wang, Yuqi Xie, Yunfan Jiang, Ajay Mandlekar, Chaowei Xiao, Yuke Zhu","year":2023,"venue":"cs.AI","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 capability and exhibits exceptional proficiency in playing Minecraft. It obtains 3.3x more unique items, travels 2.3x longer distances, and unlocks key tech tree milestones up to 15.3x faster than prior SOTA. Voyager is able to utilize the learned skill library in a new Minecraft world to solve novel tasks from scratch, while other techniques struggle to generalize. We open-source our full codebase and prompts at https://voyager.minedojo.org/.","external_url":"https://arxiv.org/abs/2305.16291","cited_by_count":null,"metadata_source":"pith","metadata_fetched_at":"2026-05-25T07:15:27.792545+00:00","pith_arxiv_id":"2305.16291","created_at":"2026-05-09T00:14:27.028890+00:00","updated_at":"2026-05-25T07:15:27.792545+00:00","title_quality_ok":true,"display_title":"Voyager: An Open-Ended Embodied Agent with Large Language Models","render_title":"Voyager: An Open-Ended Embodied Agent with Large Language Models"},"hub":{"state":{"work_id":"ffe0d207-86cf-4742-a100-e988ac8b9676","tier":"super_hub","tier_reason":"100+ Pith inbound or 10,000+ external citations","pith_inbound_count":232,"external_cited_by_count":null,"distinct_field_count":22,"first_pith_cited_at":"2023-01-10T18:12:16+00:00","last_pith_cited_at":"2026-05-22T17:59:50+00:00","author_build_status":"needed","summary_status":"needed","contexts_status":"needed","graph_status":"needed","ask_index_status":"needed","reader_status":"not_needed","recognition_status":"not_needed","updated_at":"2026-05-29T21:10:21.107801+00:00","tier_text":"super_hub"},"tier":"super_hub","role_counts":[{"context_role":"background","n":57},{"context_role":"method","n":2},{"context_role":"dataset","n":1},{"context_role":"other","n":1}],"polarity_counts":[{"context_polarity":"background","n":57},{"context_polarity":"support","n":1},{"context_polarity":"unclear","n":1},{"context_polarity":"use_dataset","n":1},{"context_polarity":"use_method","n":1}],"runs":{"ask_index":{"job_type":"ask_index","status":"succeeded","result":{"title":"Voyager: An Open-Ended Embodied Agent with Large Language Models","claims":[{"claim_text":"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 querie","claim_type":"abstract","evidence_strength":"source_metadata"}],"why_cited":"Pith tracks Voyager: An Open-Ended Embodied Agent with Large Language Models because it crossed a citation-hub threshold.","role_counts":[]},"error":null,"updated_at":"2026-05-13T23:44:00.334026+00:00"},"author_expand":{"job_type":"author_expand","status":"succeeded","result":{"authors_linked":[{"id":"757c7332-6db0-485c-8d9b-11f4b0f4b9cc","orcid":null,"display_name":"Guanzhi Wang"},{"id":"2ac69995-7333-4302-8b5e-e29b6fafe63e","orcid":null,"display_name":"Yuqi Xie"},{"id":"0915abc6-7ece-4331-9586-4998b1e7adb7","orcid":null,"display_name":"Yunfan Jiang"},{"id":"c0aff9fd-ddc7-4bb7-9631-299b0fab055a","orcid":null,"display_name":"Ajay Mandlekar"},{"id":"555e9dc7-551a-40b1-ac25-47b5c45b3f8f","orcid":null,"display_name":"Chaowei Xiao"},{"id":"5756679f-ca07-4fff-bcea-6210d51c28e4","orcid":null,"display_name":"Yuke Zhu"}]},"error":null,"updated_at":"2026-05-13T23:44:00.328813+00:00"},"context_extract":{"job_type":"context_extract","status":"succeeded","result":{"enqueued_papers":25},"error":null,"updated_at":"2026-05-13T23:34:00.770596+00:00"},"graph_features":{"job_type":"graph_features","status":"succeeded","result":{"co_cited":[{"title":"ReAct: Synergizing Reasoning and Acting in Language Models","work_id":"407a2351-25f1-497d-b611-f77d0292a8e6","shared_citers":23},{"title":"Qwen3 Technical Report","work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e","shared_citers":22},{"title":"AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation","work_id":"92b7eb9c-c3d8-4518-a376-06fa15dd895b","shared_citers":20},{"title":"DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models","work_id":"c5006563-f3ec-438a-9e35-b7b484f34828","shared_citers":19},{"title":"GPT-4 Technical Report","work_id":"b928e041-6991-4c08-8c81-0359e4097c7b","shared_citers":18},{"title":"Reflexion: Language Agents with Verbal Reinforcement Learning","work_id":"778f739e-5f55-4961-8a2a-e4736a2757f4","shared_citers":18},{"title":"A-MEM: Agentic Memory for LLM Agents","work_id":"3b98feb2-fdb1-479a-bbe4-2c298a4592e2","shared_citers":17},{"title":"Toolformer: Language Models Can Teach Themselves to Use Tools","work_id":"9bce40c8-cfd7-4983-80e0-c3bd4402322a","shared_citers":17},{"title":"MemGPT: Towards LLMs as Operating Systems","work_id":"2698f5ad-c84c-40ca-b839-0912dae10ba2","shared_citers":16},{"title":"Evaluating Large Language Models Trained on Code","work_id":"042493e9-b26f-4b4e-bbde-382072ca9b08","shared_citers":15},{"title":"WebArena: A Realistic Web Environment for Building Autonomous Agents","work_id":"7058ffd2-a339-4102-89eb-248eeb074652","shared_citers":14},{"title":"Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory","work_id":"a5aed26c-a248-48b6-a59e-f7693fcb180a","shared_citers":13},{"title":"SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning","work_id":"536a9d0a-baed-4eb9-9a51-f2b585c3388b","shared_citers":13},{"title":"SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks","work_id":"b477d12a-2ca6-4894-9f90-3fb479635e98","shared_citers":13},{"title":"ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs","work_id":"3c555b48-a4d9-42dd-9fdd-0f6018fbe9cb","shared_citers":13},{"title":"AgentBench: Evaluating LLMs as Agents","work_id":"a37549b4-4c94-412d-acc4-4efeb08509be","shared_citers":12},{"title":"Agent Skills for Large Language Models: Architecture, Acquisition, Security, and the Path Forward","work_id":"a64d3985-6826-492c-8a2e-f0965d805bfd","shared_citers":12},{"title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","work_id":"e6b75ad5-2877-4168-97c8-710407094d20","shared_citers":12},{"title":"Constitutional AI: Harmlessness from AI Feedback","work_id":"faaaa4e0-2676-4fac-a0b4-99aef10d2095","shared_citers":11},{"title":"From Local to Global: A Graph RAG Approach to Query-Focused Summarization","work_id":"588618d7-fd41-4053-b34d-a981f8793039","shared_citers":11},{"title":"Gorilla: Large Language Model Connected with Massive APIs","work_id":"126a464a-4a73-495f-b669-de1e44aa8f09","shared_citers":11},{"title":"MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents","work_id":"64baa7f3-849b-485e-b081-2074d82f1364","shared_citers":11},{"title":"MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework","work_id":"891b9780-a800-4e3c-bba0-53597ab8dc98","shared_citers":11},{"title":"Proximal Policy Optimization Algorithms","work_id":"240c67fe-d14d-4520-91c1-38a4e272ca19","shared_citers":11}],"time_series":[{"n":4,"year":2023},{"n":3,"year":2024},{"n":1,"year":2025},{"n":126,"year":2026}]},"error":null,"updated_at":"2026-05-13T23:43:54.669640+00:00"},"identity_refresh":{"job_type":"identity_refresh","status":"succeeded","result":{"fixed":1,"items":[{"title":"Qwen3 Technical Report","work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e","resolver":"local_arxiv","confidence":0.98,"old_work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e"}],"errors":[],"attempted":1},"error":null,"updated_at":"2026-05-13T23:33:56.179749+00:00"},"role_polarity":{"job_type":"role_polarity","status":"succeeded","result":{"title":"Voyager: An Open-Ended Embodied Agent with Large Language Models","claims":[{"claim_text":"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 querie","claim_type":"abstract","evidence_strength":"source_metadata"}],"why_cited":"Pith tracks Voyager: An Open-Ended Embodied Agent with Large Language Models because it crossed a citation-hub threshold.","role_counts":[]},"error":null,"updated_at":"2026-05-13T23:44:01.006997+00:00"},"summary_claims":{"job_type":"summary_claims","status":"succeeded","result":{"title":"Voyager: An Open-Ended Embodied Agent with Large Language Models","claims":[{"claim_text":"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 querie","claim_type":"abstract","evidence_strength":"source_metadata"}],"why_cited":"Pith tracks Voyager: An Open-Ended Embodied Agent with Large Language Models because it crossed a citation-hub threshold.","role_counts":[]},"error":null,"updated_at":"2026-05-13T23:44:01.010129+00:00"}},"summary":{"title":"Voyager: An Open-Ended Embodied Agent with Large Language Models","claims":[{"claim_text":"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 querie","claim_type":"abstract","evidence_strength":"source_metadata"}],"why_cited":"Pith tracks Voyager: An Open-Ended Embodied Agent with Large Language Models because it crossed a citation-hub threshold.","role_counts":[]},"graph":{"co_cited":[{"title":"ReAct: Synergizing Reasoning and Acting in Language Models","work_id":"407a2351-25f1-497d-b611-f77d0292a8e6","shared_citers":23},{"title":"Qwen3 Technical Report","work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e","shared_citers":22},{"title":"AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation","work_id":"92b7eb9c-c3d8-4518-a376-06fa15dd895b","shared_citers":20},{"title":"DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models","work_id":"c5006563-f3ec-438a-9e35-b7b484f34828","shared_citers":19},{"title":"GPT-4 Technical Report","work_id":"b928e041-6991-4c08-8c81-0359e4097c7b","shared_citers":18},{"title":"Reflexion: Language Agents with Verbal Reinforcement Learning","work_id":"778f739e-5f55-4961-8a2a-e4736a2757f4","shared_citers":18},{"title":"A-MEM: Agentic Memory for LLM Agents","work_id":"3b98feb2-fdb1-479a-bbe4-2c298a4592e2","shared_citers":17},{"title":"Toolformer: Language Models Can Teach Themselves to Use Tools","work_id":"9bce40c8-cfd7-4983-80e0-c3bd4402322a","shared_citers":17},{"title":"MemGPT: Towards LLMs as Operating Systems","work_id":"2698f5ad-c84c-40ca-b839-0912dae10ba2","shared_citers":16},{"title":"Evaluating Large Language Models Trained on Code","work_id":"042493e9-b26f-4b4e-bbde-382072ca9b08","shared_citers":15},{"title":"WebArena: A Realistic Web Environment for Building Autonomous Agents","work_id":"7058ffd2-a339-4102-89eb-248eeb074652","shared_citers":14},{"title":"Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory","work_id":"a5aed26c-a248-48b6-a59e-f7693fcb180a","shared_citers":13},{"title":"SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning","work_id":"536a9d0a-baed-4eb9-9a51-f2b585c3388b","shared_citers":13},{"title":"SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks","work_id":"b477d12a-2ca6-4894-9f90-3fb479635e98","shared_citers":13},{"title":"ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs","work_id":"3c555b48-a4d9-42dd-9fdd-0f6018fbe9cb","shared_citers":13},{"title":"AgentBench: Evaluating LLMs as Agents","work_id":"a37549b4-4c94-412d-acc4-4efeb08509be","shared_citers":12},{"title":"Agent Skills for Large Language Models: Architecture, Acquisition, Security, and the Path Forward","work_id":"a64d3985-6826-492c-8a2e-f0965d805bfd","shared_citers":12},{"title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","work_id":"e6b75ad5-2877-4168-97c8-710407094d20","shared_citers":12},{"title":"Constitutional AI: Harmlessness from AI Feedback","work_id":"faaaa4e0-2676-4fac-a0b4-99aef10d2095","shared_citers":11},{"title":"From Local to Global: A Graph RAG Approach to Query-Focused Summarization","work_id":"588618d7-fd41-4053-b34d-a981f8793039","shared_citers":11},{"title":"Gorilla: Large Language Model Connected with Massive APIs","work_id":"126a464a-4a73-495f-b669-de1e44aa8f09","shared_citers":11},{"title":"MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents","work_id":"64baa7f3-849b-485e-b081-2074d82f1364","shared_citers":11},{"title":"MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework","work_id":"891b9780-a800-4e3c-bba0-53597ab8dc98","shared_citers":11},{"title":"Proximal Policy Optimization Algorithms","work_id":"240c67fe-d14d-4520-91c1-38a4e272ca19","shared_citers":11}],"time_series":[{"n":4,"year":2023},{"n":3,"year":2024},{"n":1,"year":2025},{"n":126,"year":2026}]},"authors":[{"id":"c0aff9fd-ddc7-4bb7-9631-299b0fab055a","orcid":null,"display_name":"Ajay Mandlekar","source":"manual","import_confidence":0.72},{"id":"555e9dc7-551a-40b1-ac25-47b5c45b3f8f","orcid":null,"display_name":"Chaowei Xiao","source":"manual","import_confidence":0.72},{"id":"757c7332-6db0-485c-8d9b-11f4b0f4b9cc","orcid":null,"display_name":"Guanzhi Wang","source":"manual","import_confidence":0.72},{"id":"5756679f-ca07-4fff-bcea-6210d51c28e4","orcid":null,"display_name":"Yuke Zhu","source":"manual","import_confidence":0.72},{"id":"0915abc6-7ece-4331-9586-4998b1e7adb7","orcid":null,"display_name":"Yunfan Jiang","source":"manual","import_confidence":0.72},{"id":"2ac69995-7333-4302-8b5e-e29b6fafe63e","orcid":null,"display_name":"Yuqi Xie","source":"manual","import_confidence":0.72}]}}