{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NU6IWKEYUV475VS32XCFWE3HUV","short_pith_number":"pith:NU6IWKEY","schema_version":"1.0","canonical_sha256":"6d3c8b2898a579fed65bd5c45b1367a5531da27ae7866213aa908ab59448581f","source":{"kind":"arxiv","id":"2512.02473","version":2},"attestation_state":"computed","paper":{"title":"WorldPack: Dynamic Frame Compression for Long-context Video World Modeling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Hiroki Furuta, Masahiro Suzuki, Yusuke Iwasawa, Yutaka Matsuo, Yuta Oshima","submitted_at":"2025-12-02T07:06:23Z","abstract_excerpt":"Video world models have attracted significant attention for their ability to produce high-fidelity future visual observations conditioned on past observations and navigation actions. However, achieving temporally and spatially consistent generation over long horizons remains an open challenge: existing approaches either compress past frames using generic importance schedules that do not exploit 3D viewpoint geometry, or retrieve only a handful of spatially relevant frames without increasing the total amount of retained history. In this paper, we propose WorldPack, a video world model that intr"},"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":"2512.02473","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-12-02T07:06:23Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"f800644e22f3ead94f1e3edde67aabb42b035537ff46c0e8ad38d9027faf08fd","abstract_canon_sha256":"ef463d60499c2b431f159ece9dfa305f9b91a09704ec59abd3bbdb0528bd86eb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T01:24:18.763778Z","signature_b64":"NR8Nvy7GnoQaAY0Gr82VPrVmH7FZlypNM/zJxAcJz0PommIfF0V6k0bN4OOiYQ6Fkfd/gacEAkR4TJcMTqesCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6d3c8b2898a579fed65bd5c45b1367a5531da27ae7866213aa908ab59448581f","last_reissued_at":"2026-07-23T01:24:18.762852Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T01:24:18.762852Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"WorldPack: Dynamic Frame Compression for Long-context Video World Modeling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Hiroki Furuta, Masahiro Suzuki, Yusuke Iwasawa, Yutaka Matsuo, Yuta Oshima","submitted_at":"2025-12-02T07:06:23Z","abstract_excerpt":"Video world models have attracted significant attention for their ability to produce high-fidelity future visual observations conditioned on past observations and navigation actions. However, achieving temporally and spatially consistent generation over long horizons remains an open challenge: existing approaches either compress past frames using generic importance schedules that do not exploit 3D viewpoint geometry, or retrieve only a handful of spatially relevant frames without increasing the total amount of retained history. In this paper, we propose WorldPack, a video world model that intr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.02473","kind":"arxiv","version":2},"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/2512.02473/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":"2512.02473","created_at":"2026-07-23T01:24:18.763281+00:00"},{"alias_kind":"arxiv_version","alias_value":"2512.02473v2","created_at":"2026-07-23T01:24:18.763281+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.02473","created_at":"2026-07-23T01:24:18.763281+00:00"},{"alias_kind":"pith_short_12","alias_value":"NU6IWKEYUV47","created_at":"2026-07-23T01:24:18.763281+00:00"},{"alias_kind":"pith_short_16","alias_value":"NU6IWKEYUV475VS3","created_at":"2026-07-23T01:24:18.763281+00:00"},{"alias_kind":"pith_short_8","alias_value":"NU6IWKEY","created_at":"2026-07-23T01:24:18.763281+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":4,"sample":[{"citing_arxiv_id":"2606.02436","citing_title":"Geometry-Aware Implicit Memory for Video World Models","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2605.30855","citing_title":"Robust Dreamer: Deviation-Aware Latent Gaussian Memory for Action-Controlled AR Video Generation","ref_index":42,"is_internal_anchor":true},{"citing_arxiv_id":"2603.28489","citing_title":"Video Generation Models as World Models: Efficient Paradigms, Architectures and Algorithms","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11367","citing_title":"3D-Belief: Embodied Belief Inference via Generative 3D World Modeling","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV","json":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV.json","graph_json":"https://pith.science/api/pith-number/NU6IWKEYUV475VS32XCFWE3HUV/graph.json","events_json":"https://pith.science/api/pith-number/NU6IWKEYUV475VS32XCFWE3HUV/events.json","paper":"https://pith.science/paper/NU6IWKEY"},"agent_actions":{"view_html":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV","download_json":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV.json","view_paper":"https://pith.science/paper/NU6IWKEY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2512.02473&json=true","fetch_graph":"https://pith.science/api/pith-number/NU6IWKEYUV475VS32XCFWE3HUV/graph.json","fetch_events":"https://pith.science/api/pith-number/NU6IWKEYUV475VS32XCFWE3HUV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV/action/storage_attestation","attest_author":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV/action/author_attestation","sign_citation":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV/action/citation_signature","submit_replication":"https://pith.science/pith/NU6IWKEYUV475VS32XCFWE3HUV/action/replication_record"}},"created_at":"2026-07-23T01:24:18.763281+00:00","updated_at":"2026-07-23T01:24:18.763281+00:00"}