{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:2KQAC6XRQOM6TWTRAK775KA5MJ","short_pith_number":"pith:2KQAC6XR","schema_version":"1.0","canonical_sha256":"d2a0017af18399e9da7102bffea81d624a64a97bcd1249fd2c2292175ef73a3a","source":{"kind":"arxiv","id":"2211.10831","version":1},"attestation_state":"computed","paper":{"title":"Joint Embedding Predictive Architectures Focus on Slow Features","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jyothir S V, Kyunghyun Cho, Nicolas Carion, Siddhartha Jalagam, Vlad Sobal, Yann LeCun","submitted_at":"2022-11-20T00:50:11Z","abstract_excerpt":"Many common methods for learning a world model for pixel-based environments use generative architectures trained with pixel-level reconstruction objectives. Recently proposed Joint Embedding Predictive Architectures (JEPA) offer a reconstruction-free alternative. In this work, we analyze performance of JEPA trained with VICReg and SimCLR objectives in the fully offline setting without access to rewards, and compare the results to the performance of the generative architecture. We test the methods in a simple environment with a moving dot with various background distractors, and probe learned r"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2211.10831","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-11-20T00:50:11Z","cross_cats_sorted":[],"title_canon_sha256":"3bdf0f0c7673a8137ba4849506330b56030422e84bb6598f6a94a5304c516008","abstract_canon_sha256":"1a39e54cd16fa61b849adc8384fd8d844d0eb184e3315cbbedd1b82657994620"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:17:35.322336Z","signature_b64":"5D5foBtWJUAolrI0mnFFVr4NdvBS1uIG1xAEoIKRzlnUEzsdlNQHAf+fhKbiy0dUzvOnVQN/GNqSH88Bml0lAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d2a0017af18399e9da7102bffea81d624a64a97bcd1249fd2c2292175ef73a3a","last_reissued_at":"2026-07-05T05:17:35.321785Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:17:35.321785Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Joint Embedding Predictive Architectures Focus on Slow Features","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Jyothir S V, Kyunghyun Cho, Nicolas Carion, Siddhartha Jalagam, Vlad Sobal, Yann LeCun","submitted_at":"2022-11-20T00:50:11Z","abstract_excerpt":"Many common methods for learning a world model for pixel-based environments use generative architectures trained with pixel-level reconstruction objectives. Recently proposed Joint Embedding Predictive Architectures (JEPA) offer a reconstruction-free alternative. In this work, we analyze performance of JEPA trained with VICReg and SimCLR objectives in the fully offline setting without access to rewards, and compare the results to the performance of the generative architecture. We test the methods in a simple environment with a moving dot with various background distractors, and probe learned r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.10831","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2211.10831/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2211.10831","created_at":"2026-07-05T05:17:35.321856+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.10831v1","created_at":"2026-07-05T05:17:35.321856+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.10831","created_at":"2026-07-05T05:17:35.321856+00:00"},{"alias_kind":"pith_short_12","alias_value":"2KQAC6XRQOM6","created_at":"2026-07-05T05:17:35.321856+00:00"},{"alias_kind":"pith_short_16","alias_value":"2KQAC6XRQOM6TWTR","created_at":"2026-07-05T05:17:35.321856+00:00"},{"alias_kind":"pith_short_8","alias_value":"2KQAC6XR","created_at":"2026-07-05T05:17:35.321856+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26217","citing_title":"Fast LeWorldModel","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07365","citing_title":"A robust PPG foundation model using multimodal physiological supervision","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07770","citing_title":"Contrast encodes inductive bias: separating slow noise from dynamics in predictive representation learning","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26379","citing_title":"When Does LeJEPA Learn a World Model?","ref_index":73,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08732","citing_title":"Latent Geometry Beyond Search: Amortizing Planning in World Models","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2603.19312","citing_title":"LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08732","citing_title":"Latent Geometry Beyond Search: Amortizing Planning in World Models","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05586","citing_title":"AeroJEPA: Learning Semantic Latent Representations for Scalable 3D Aerodynamic Field Modeling","ref_index":28,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ","json":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ.json","graph_json":"https://pith.science/api/pith-number/2KQAC6XRQOM6TWTRAK775KA5MJ/graph.json","events_json":"https://pith.science/api/pith-number/2KQAC6XRQOM6TWTRAK775KA5MJ/events.json","paper":"https://pith.science/paper/2KQAC6XR"},"agent_actions":{"view_html":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ","download_json":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ.json","view_paper":"https://pith.science/paper/2KQAC6XR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.10831&json=true","fetch_graph":"https://pith.science/api/pith-number/2KQAC6XRQOM6TWTRAK775KA5MJ/graph.json","fetch_events":"https://pith.science/api/pith-number/2KQAC6XRQOM6TWTRAK775KA5MJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ/action/storage_attestation","attest_author":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ/action/author_attestation","sign_citation":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ/action/citation_signature","submit_replication":"https://pith.science/pith/2KQAC6XRQOM6TWTRAK775KA5MJ/action/replication_record"}},"created_at":"2026-07-05T05:17:35.321856+00:00","updated_at":"2026-07-05T05:17:35.321856+00:00"}