{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/arxiv-2603-04553","title":"Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling","arxiv_id":"2603.04553","date":"2026-03-04","proceeding":null,"authors":["Tal Daniel","Carl Qi","Dan Haramati","Amir Zadeh","Chuan Li","Aviv Tamar","Deepak Pathak","David Held"],"abstract":"We introduce Latent Particle World Model (LPWM), a self-supervised object-centric world model scaled to real-world multi-object datasets and applicable in decision-making. LPWM autonomously discovers keypoints, bounding boxes, and object masks directly from video data, enabling it to learn rich scene decompositions without supervision. Our architecture is trained end-to-end purely from videos and supports flexible conditioning on actions, language, and image goals. LPWM models stochastic particle dynamics via a novel latent action module and achieves state-of-the-art results on diverse real-world and synthetic datasets. Beyond stochastic video modeling, LPWM is readily applicable to decision-making, including goal-conditioned imitation learning, as we demonstrate in the paper. Code, data, pre-trained models and video rollouts are available: https://taldatech.github.io/lpwm-web","url_abs":"https://arxiv.org/abs/2603.04553","url_pdf":"https://arxiv.org/pdf/2603.04553","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2603.04553","atlas_url":"https://app.syntology.ai/?focus=2603.04553","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2603.04553"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/taldatech/lpwm","reach":null}],"summary":{"unverified":8},"by_repo_kind":{"found_in_text":{"samples":8,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"359375634e1bfb04","entry":"batch_pairwise_dist","repo":"taldatech/lpwm","repo_kind":"found_in_text","path":"utils/loss_functions.py","file_url":"https://github.com/taldatech/lpwm/blob/HEAD/utils/loss_functions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"359375634e1bfb04"}},{"code_sha256_prefix":"2449ff1e77071de5","entry":"batch_pairwise_kl","repo":"taldatech/lpwm","repo_kind":"found_in_text","path":"utils/loss_functions.py","file_url":"https://github.com/taldatech/lpwm/blob/HEAD/utils/loss_functions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2449ff1e77071de5"}},{"code_sha256_prefix":"ed26747362a660ae","entry":"calc_kl_jit","repo":"taldatech/lpwm","repo_kind":"found_in_text","path":"models.py","file_url":"https://github.com/taldatech/lpwm/blob/HEAD/models.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ed26747362a660ae"}},{"code_sha256_prefix":"a1be5175025a5488","entry":"calc_model_size","repo":"taldatech/lpwm","repo_kind":"found_in_text","path":"modules/vision_modules.py","file_url":"https://github.com/taldatech/lpwm/blob/HEAD/modules/vision_modules.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a1be5175025a5488"}},{"code_sha256_prefix":"c48a86b29b5c0a69","entry":"calc_reconstruction_loss","repo":"taldatech/lpwm","repo_kind":"found_in_text","path":"utils/loss_functions.py","file_url":"https://github.com/taldatech/lpwm/blob/HEAD/utils/loss_functions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c48a86b29b5c0a69"}},{"code_sha256_prefix":"c8d9d38b144e401f","entry":"md5_hash","repo":"taldatech/lpwm","repo_kind":"found_in_text","path":"modules/vision_modules.py","file_url":"https://github.com/taldatech/lpwm/blob/HEAD/modules/vision_modules.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c8d9d38b144e401f"}},{"code_sha256_prefix":"7ba4bcfedd050bf0","entry":"nonlinearity","repo":"taldatech/lpwm","repo_kind":"found_in_text","path":"modules/vision_modules.py","file_url":"https://github.com/taldatech/lpwm/blob/HEAD/modules/vision_modules.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7ba4bcfedd050bf0"}},{"code_sha256_prefix":"cbdb2f43d720b713","entry":"reparam","repo":"taldatech/lpwm","repo_kind":"found_in_text","path":"models.py","file_url":"https://github.com/taldatech/lpwm/blob/HEAD/models.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"cbdb2f43d720b713"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}