{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7CBP4H6CR4Q66SASLJXPVOD73R","short_pith_number":"pith:7CBP4H6C","schema_version":"1.0","canonical_sha256":"f882fe1fc28f21ef48125a6efab87fdc636dc17b03e399442db65736f838a33c","source":{"kind":"arxiv","id":"2606.31232","version":1},"attestation_state":"computed","paper":{"title":"Delta-JEPA: Learning Action-Sensitive World Models via Latent Difference Decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bingkang Shi, Chenxi Bao, Guoqing Chao, Hongzhu Yi, Jingjing Zhou, Jungang Xu, Tao Yu, Tiankun Yang, Tianyu Zong, Xingchen Chen, Yuanxiang Wang, Yujia Yang, Zhenghao Zhang, Zhenyu Guan","submitted_at":"2026-06-30T07:08:24Z","abstract_excerpt":"Learning visual world models for planning requires compact latent dynamics that remain sensitive to actions, yet reconstruction-free joint-embedding objectives can collapse to action-insensitive representations. We propose Delta-JEPA, an end-to-end reconstruction-free world model that augments latent forward prediction with a Latent Difference Action Decoder (LDAD). Unlike inverse decoders that infer actions from concatenated endpoint embeddings, LDAD reconstructs the executed action from the latent displacement between consecutive observations. This displacement-level supervision directly reg"},"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":"2606.31232","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-30T07:08:24Z","cross_cats_sorted":[],"title_canon_sha256":"ba1448be03e44f44b252d33eeb9a2d03a7a76126f886ceb407f5b274a8fa3c10","abstract_canon_sha256":"8b9c073b35411f25190e78d3393c6ad3f6ff3f14b7c5eaf4ab57762a4b0e7a87"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-01T01:17:56.484585Z","signature_b64":"LO+9q5gMOQCo+mKErenUDwvEeV0Rq/lOphhCgJmx0BfErKutbFEZXKn/4exFd454q6lQAg9Mlz7iGM3GGgInCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f882fe1fc28f21ef48125a6efab87fdc636dc17b03e399442db65736f838a33c","last_reissued_at":"2026-07-01T01:17:56.483939Z","signature_status":"signed_v1","first_computed_at":"2026-07-01T01:17:56.483939Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Delta-JEPA: Learning Action-Sensitive World Models via Latent Difference Decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Bingkang Shi, Chenxi Bao, Guoqing Chao, Hongzhu Yi, Jingjing Zhou, Jungang Xu, Tao Yu, Tiankun Yang, Tianyu Zong, Xingchen Chen, Yuanxiang Wang, Yujia Yang, Zhenghao Zhang, Zhenyu Guan","submitted_at":"2026-06-30T07:08:24Z","abstract_excerpt":"Learning visual world models for planning requires compact latent dynamics that remain sensitive to actions, yet reconstruction-free joint-embedding objectives can collapse to action-insensitive representations. We propose Delta-JEPA, an end-to-end reconstruction-free world model that augments latent forward prediction with a Latent Difference Action Decoder (LDAD). Unlike inverse decoders that infer actions from concatenated endpoint embeddings, LDAD reconstructs the executed action from the latent displacement between consecutive observations. This displacement-level supervision directly reg"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.31232","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/2606.31232/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":"2606.31232","created_at":"2026-07-01T01:17:56.484014+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.31232v1","created_at":"2026-07-01T01:17:56.484014+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.31232","created_at":"2026-07-01T01:17:56.484014+00:00"},{"alias_kind":"pith_short_12","alias_value":"7CBP4H6CR4Q6","created_at":"2026-07-01T01:17:56.484014+00:00"},{"alias_kind":"pith_short_16","alias_value":"7CBP4H6CR4Q66SAS","created_at":"2026-07-01T01:17:56.484014+00:00"},{"alias_kind":"pith_short_8","alias_value":"7CBP4H6C","created_at":"2026-07-01T01:17:56.484014+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R","json":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R.json","graph_json":"https://pith.science/api/pith-number/7CBP4H6CR4Q66SASLJXPVOD73R/graph.json","events_json":"https://pith.science/api/pith-number/7CBP4H6CR4Q66SASLJXPVOD73R/events.json","paper":"https://pith.science/paper/7CBP4H6C"},"agent_actions":{"view_html":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R","download_json":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R.json","view_paper":"https://pith.science/paper/7CBP4H6C","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.31232&json=true","fetch_graph":"https://pith.science/api/pith-number/7CBP4H6CR4Q66SASLJXPVOD73R/graph.json","fetch_events":"https://pith.science/api/pith-number/7CBP4H6CR4Q66SASLJXPVOD73R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R/action/storage_attestation","attest_author":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R/action/author_attestation","sign_citation":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R/action/citation_signature","submit_replication":"https://pith.science/pith/7CBP4H6CR4Q66SASLJXPVOD73R/action/replication_record"}},"created_at":"2026-07-01T01:17:56.484014+00:00","updated_at":"2026-07-01T01:17:56.484014+00:00"}