{"work":{"id":"9a85fc69-74df-450e-94cd-69d186e9e830","openalex_id":null,"doi":null,"arxiv_id":"2602.15922","raw_key":null,"title":"World Action Models are Zero-shot Policies","authors":null,"authors_text":"Seonghyeon Ye, Yunhao Ge, Kaiyuan Zheng, Shenyuan Gao, Sihyun Yu, George Kurian","year":2026,"venue":"cs.RO","abstract":"State-of-the-art Vision-Language-Action (VLA) models excel at semantic generalization but struggle to generalize to unseen physical motions in novel environments. We introduce DreamZero, a World Action Model (WAM) built upon a pretrained video diffusion backbone. Unlike VLAs, WAMs learn physical dynamics by predicting future world states and actions, using video as a dense representation of how the world evolves. By jointly modeling video and action, DreamZero learns diverse skills effectively from heterogeneous robot data without relying on repetitive demonstrations. This results in over 2x improvement in generalization to new tasks and environments compared to state-of-the-art VLAs in real robot experiments. Crucially, through model and system optimizations, we enable a 14B autoregressive video diffusion model to perform real-time closed-loop control at 7Hz. Finally, we demonstrate two forms of cross-embodiment transfer: video-only demonstrations from other robots or humans yield a relative improvement of over 42% on unseen task performance with just 10-20 minutes of data. More surprisingly, DreamZero enables few-shot embodiment adaptation, transferring to a new embodiment with only 30 minutes of play data while retaining zero-shot generalization.","external_url":"https://arxiv.org/abs/2602.15922","cited_by_count":null,"metadata_source":"pith","metadata_fetched_at":"2026-05-25T03:56:37.117120+00:00","pith_arxiv_id":"2602.15922","created_at":"2026-05-09T06:05:34.936500+00:00","updated_at":"2026-06-05T21:23:00.469572+00:00","title_quality_ok":true,"display_title":"World Action Models are Zero-shot Policies","render_title":"World Action Models are Zero-shot Policies"},"hub":{"state":{"work_id":"9a85fc69-74df-450e-94cd-69d186e9e830","tier":"hub","tier_reason":"10+ Pith inbound or 1,000+ external citations","pith_inbound_count":56,"external_cited_by_count":null,"distinct_field_count":5,"first_pith_cited_at":"2026-01-11T21:00:58+00:00","last_pith_cited_at":"2026-05-22T17:08:37+00:00","author_build_status":"not_needed","summary_status":"needed","contexts_status":"needed","graph_status":"needed","ask_index_status":"not_needed","reader_status":"not_needed","recognition_status":"not_needed","updated_at":"2026-06-10T23:37:18.941190+00:00","tier_text":"hub"},"tier":"hub","role_counts":[{"context_role":"background","n":28},{"context_role":"baseline","n":1}],"polarity_counts":[{"context_polarity":"background","n":26},{"context_polarity":"unclear","n":2},{"context_polarity":"baseline","n":1}],"runs":{"context_extract":{"job_type":"context_extract","status":"succeeded","result":{"enqueued_papers":25},"error":null,"updated_at":"2026-05-14T17:59:31.376222+00:00"},"graph_features":{"job_type":"graph_features","status":"succeeded","result":{"co_cited":[{"title":"$\\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization","work_id":"d1ad7304-d09a-49bc-809e-846439f6aff9","shared_citers":23},{"title":"$\\pi_0$: A Vision-Language-Action Flow Model for General Robot Control","work_id":"f790abdc-a796-482f-a40d-f8ee035ecfc2","shared_citers":22},{"title":"Wan: Open and Advanced Large-Scale Video Generative Models","work_id":"ad3ebc3b-4224-46c9-b61d-bcf135da0a7c","shared_citers":22},{"title":"Causal World Modeling for Robot Control","work_id":"a33c4ee0-db06-4f9a-8852-c62e3a72fc27","shared_citers":21},{"title":"OpenVLA: An Open-Source Vision-Language-Action Model","work_id":"3e7e65c5-5aed-4fe9-8414-2092bcb31cc7","shared_citers":21},{"title":"Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning","work_id":"3d63039f-41b0-4a31-af31-6fc10f5c1b1b","shared_citers":17},{"title":"Motus: A Unified Latent Action World Model","work_id":"d0b2d257-524d-4d67-9daf-5fb43e5e977a","shared_citers":16},{"title":"Fast-WAM: Do World Action Models Need Test-time Future Imagination?","work_id":"772d0226-9ad6-42c0-9c79-d36d37edbbe4","shared_citers":15},{"title":"GR00T N1: An Open Foundation Model for Generalist Humanoid Robots","work_id":"e2db69c7-ee8a-4cb7-a761-7b8de1dfcf97","shared_citers":15},{"title":"GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulation","work_id":"843ab5eb-2815-4db8-b3bc-890b23fa5ffa","shared_citers":13},{"title":"Unified Video Action Model","work_id":"fb4cc512-d1d9-40f4-8854-d35950ad20b3","shared_citers":13},{"title":"Flow Matching for Generative Modeling","work_id":"6edb71c4-5d64-40af-a394-9757ea051a36","shared_citers":12},{"title":"RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation","work_id":"9b985126-4a2f-4bdf-b014-2a7524ec634e","shared_citers":12},{"title":"Video generators are robot policies","work_id":"0941f8fe-e893-4e3a-9ac6-0a72f26340e9","shared_citers":12},{"title":"Video Prediction Policy: A Generalist Robot Policy with Predictive Visual Representations","work_id":"62dbe235-8473-4190-8686-17e7437de50f","shared_citers":12},{"title":"Unified World Models: Coupling Video and Action Diffusion for Pretraining on Large Robotic Datasets","work_id":"c181dcf1-e774-4216-a30e-e55c3f3a766c","shared_citers":11},{"title":"AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systems","work_id":"f797e9ec-510f-43a7-8a0c-18009ce332e5","shared_citers":10},{"title":"Cosmos World Foundation Model Platform for Physical AI","work_id":"a2dba24c-318d-476a-8b21-4289c265810c","shared_citers":10},{"title":"mimic-video: Video-action models for generalizable robot control beyond vlas.arXiv preprint 2512.15692","work_id":"cd5b191a-8f67-43d3-8816-0ade1b9a7c29","shared_citers":10},{"title":"RT-1: Robotics Transformer for Real-World Control at Scale","work_id":"e11bda85-8531-46bc-a07f-d0ade3643ab1","shared_citers":10},{"title":"SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics","work_id":"0c5e9314-5fa7-4613-ad12-605a71d561d2","shared_citers":9},{"title":"SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Model","work_id":"592041b3-3ca2-4836-8dd4-f8095d8a692b","shared_citers":9},{"title":"$\\pi^{*}_{0.6}$: a VLA That Learns From Experience","work_id":"7c1b3355-694a-44c6-880f-631e897e1713","shared_citers":8},{"title":"Ctrl-world: A controllable generative world model for robot manipulation","work_id":"60c6a353-4bc6-4c87-8c39-372fcddd6ac0","shared_citers":8}],"time_series":[{"n":38,"year":2026}],"dependency_candidates":[]},"error":null,"updated_at":"2026-05-14T17:59:48.243853+00:00"},"identity_refresh":{"job_type":"identity_refresh","status":"succeeded","result":{"items":[{"title":"Qwen3 Technical Report","outcome":"unchanged","work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e","resolver":"local_arxiv","confidence":0.98,"old_work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e"}],"counts":{"fixed":0,"merged":0,"unchanged":1,"quarantined":0,"needs_external_resolution":0},"errors":[],"attempted":1},"error":null,"updated_at":"2026-05-14T17:59:36.590343+00:00"},"summary_claims":{"job_type":"summary_claims","status":"succeeded","result":{"title":"World Action Models are Zero-shot Policies","claims":[{"claim_text":"State-of-the-art Vision-Language-Action (VLA) models excel at semantic generalization but struggle to generalize to unseen physical motions in novel environments. We introduce DreamZero, a World Action Model (WAM) built upon a pretrained video diffusion backbone. Unlike VLAs, WAMs learn physical dynamics by predicting future world states and actions, using video as a dense representation of how the world evolves. By jointly modeling video and action, DreamZero learns diverse skills effectively from heterogeneous robot data without relying on repetitive demonstrations. This results in over 2x i","claim_type":"abstract","evidence_strength":"source_metadata"}],"why_cited":"Pith tracks World Action Models are Zero-shot Policies because it crossed a citation-hub threshold.","role_counts":[]},"error":null,"updated_at":"2026-05-14T18:00:32.605740+00:00"}},"summary":{"title":"World Action Models are Zero-shot Policies","claims":[{"claim_text":"State-of-the-art Vision-Language-Action (VLA) models excel at semantic generalization but struggle to generalize to unseen physical motions in novel environments. We introduce DreamZero, a World Action Model (WAM) built upon a pretrained video diffusion backbone. Unlike VLAs, WAMs learn physical dynamics by predicting future world states and actions, using video as a dense representation of how the world evolves. By jointly modeling video and action, DreamZero learns diverse skills effectively from heterogeneous robot data without relying on repetitive demonstrations. This results in over 2x i","claim_type":"abstract","evidence_strength":"source_metadata"}],"why_cited":"Pith tracks World Action Models are Zero-shot Policies because it crossed a citation-hub threshold.","role_counts":[]},"graph":{"co_cited":[{"title":"$\\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization","work_id":"d1ad7304-d09a-49bc-809e-846439f6aff9","shared_citers":23},{"title":"$\\pi_0$: A Vision-Language-Action Flow Model for General Robot Control","work_id":"f790abdc-a796-482f-a40d-f8ee035ecfc2","shared_citers":22},{"title":"Wan: Open and Advanced Large-Scale Video Generative Models","work_id":"ad3ebc3b-4224-46c9-b61d-bcf135da0a7c","shared_citers":22},{"title":"Causal World Modeling for Robot Control","work_id":"a33c4ee0-db06-4f9a-8852-c62e3a72fc27","shared_citers":21},{"title":"OpenVLA: An Open-Source Vision-Language-Action Model","work_id":"3e7e65c5-5aed-4fe9-8414-2092bcb31cc7","shared_citers":21},{"title":"Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning","work_id":"3d63039f-41b0-4a31-af31-6fc10f5c1b1b","shared_citers":17},{"title":"Motus: A Unified Latent Action World Model","work_id":"d0b2d257-524d-4d67-9daf-5fb43e5e977a","shared_citers":16},{"title":"Fast-WAM: Do World Action Models Need Test-time Future Imagination?","work_id":"772d0226-9ad6-42c0-9c79-d36d37edbbe4","shared_citers":15},{"title":"GR00T N1: An Open Foundation Model for Generalist Humanoid Robots","work_id":"e2db69c7-ee8a-4cb7-a761-7b8de1dfcf97","shared_citers":15},{"title":"GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulation","work_id":"843ab5eb-2815-4db8-b3bc-890b23fa5ffa","shared_citers":13},{"title":"Unified Video Action Model","work_id":"fb4cc512-d1d9-40f4-8854-d35950ad20b3","shared_citers":13},{"title":"Flow Matching for Generative Modeling","work_id":"6edb71c4-5d64-40af-a394-9757ea051a36","shared_citers":12},{"title":"RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation","work_id":"9b985126-4a2f-4bdf-b014-2a7524ec634e","shared_citers":12},{"title":"Video generators are robot policies","work_id":"0941f8fe-e893-4e3a-9ac6-0a72f26340e9","shared_citers":12},{"title":"Video Prediction Policy: A Generalist Robot Policy with Predictive Visual Representations","work_id":"62dbe235-8473-4190-8686-17e7437de50f","shared_citers":12},{"title":"Unified World Models: Coupling Video and Action Diffusion for Pretraining on Large Robotic Datasets","work_id":"c181dcf1-e774-4216-a30e-e55c3f3a766c","shared_citers":11},{"title":"AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systems","work_id":"f797e9ec-510f-43a7-8a0c-18009ce332e5","shared_citers":10},{"title":"Cosmos World Foundation Model Platform for Physical AI","work_id":"a2dba24c-318d-476a-8b21-4289c265810c","shared_citers":10},{"title":"mimic-video: Video-action models for generalizable robot control beyond vlas.arXiv preprint 2512.15692","work_id":"cd5b191a-8f67-43d3-8816-0ade1b9a7c29","shared_citers":10},{"title":"RT-1: Robotics Transformer for Real-World Control at Scale","work_id":"e11bda85-8531-46bc-a07f-d0ade3643ab1","shared_citers":10},{"title":"SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics","work_id":"0c5e9314-5fa7-4613-ad12-605a71d561d2","shared_citers":9},{"title":"SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Model","work_id":"592041b3-3ca2-4836-8dd4-f8095d8a692b","shared_citers":9},{"title":"$\\pi^{*}_{0.6}$: a VLA That Learns From Experience","work_id":"7c1b3355-694a-44c6-880f-631e897e1713","shared_citers":8},{"title":"Ctrl-world: A controllable generative world model for robot manipulation","work_id":"60c6a353-4bc6-4c87-8c39-372fcddd6ac0","shared_citers":8}],"time_series":[{"n":38,"year":2026}],"dependency_candidates":[]},"authors":[]}}