{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AQ4KYZHHINCEEMOE7PLBMWJOOF","short_pith_number":"pith:AQ4KYZHH","schema_version":"1.0","canonical_sha256":"0438ac64e743444231c4fbd616592e7140f2a143745be5e9814610d321d0083c","source":{"kind":"arxiv","id":"2506.03141","version":2},"attestation_state":"computed","paper":{"title":"Context as Memory: Scene-Consistent Interactive Long Video Generation with Memory Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Di Zhang, Jianhong Bai, Jiwen Yu, Pengfei Wan, Quande Liu, Xihui Liu, Xintao Wang, Yiran Qin","submitted_at":"2025-06-03T17:59:05Z","abstract_excerpt":"Recent advances in interactive video generation have shown promising results, yet existing approaches struggle with scene-consistent memory capabilities in long video generation due to limited use of historical context. In this work, we propose Context-as-Memory, which utilizes historical context as memory for video generation. It includes two simple yet effective designs: (1) storing context in frame format without additional post-processing; (2) conditioning by concatenating context and frames to be predicted along the frame dimension at the input, requiring no external control modules. Furt"},"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":"2506.03141","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-03T17:59:05Z","cross_cats_sorted":[],"title_canon_sha256":"84eb25036f0757903a05c382f25761591858a269cc868796a7c0113738fd777d","abstract_canon_sha256":"b537542e47f1d7e8080edbaeffc82630116bcce8ef31bd2a498efab199dcd552"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:52:20.843102Z","signature_b64":"brctaxHxrLX4bjpkX82S7wTaDUrtpJVyTUO6AUaoZgq//cGDWkjS21pgzvyTcfW6Yt/7PsBA94gldKFhQcbsAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0438ac64e743444231c4fbd616592e7140f2a143745be5e9814610d321d0083c","last_reissued_at":"2026-07-05T11:52:20.842589Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:52:20.842589Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Context as Memory: Scene-Consistent Interactive Long Video Generation with Memory Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Di Zhang, Jianhong Bai, Jiwen Yu, Pengfei Wan, Quande Liu, Xihui Liu, Xintao Wang, Yiran Qin","submitted_at":"2025-06-03T17:59:05Z","abstract_excerpt":"Recent advances in interactive video generation have shown promising results, yet existing approaches struggle with scene-consistent memory capabilities in long video generation due to limited use of historical context. In this work, we propose Context-as-Memory, which utilizes historical context as memory for video generation. It includes two simple yet effective designs: (1) storing context in frame format without additional post-processing; (2) conditioning by concatenating context and frames to be predicted along the frame dimension at the input, requiring no external control modules. Furt"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.03141","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/2506.03141/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":"2506.03141","created_at":"2026-07-05T11:52:20.842648+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.03141v2","created_at":"2026-07-05T11:52:20.842648+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.03141","created_at":"2026-07-05T11:52:20.842648+00:00"},{"alias_kind":"pith_short_12","alias_value":"AQ4KYZHHINCE","created_at":"2026-07-05T11:52:20.842648+00:00"},{"alias_kind":"pith_short_16","alias_value":"AQ4KYZHHINCEEMOE","created_at":"2026-07-05T11:52:20.842648+00:00"},{"alias_kind":"pith_short_8","alias_value":"AQ4KYZHH","created_at":"2026-07-05T11:52:20.842648+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":18,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.16449","citing_title":"PermaVid: Consistent Video Generation Across Edits via Disentangled Context Memory","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02517","citing_title":"WorldDirector: Building Controllable World Simulators with Persistent Dynamic Memory","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10671","citing_title":"FadeMem: Distance-Aware Memory Consolidation for Autoregressive Video Diffusion","ref_index":82,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09828","citing_title":"Latent Spatial Memory for Video World Models","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09803","citing_title":"Echo-Memory: A Controlled Study of Memory in Action World Models","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02753","citing_title":"MetaWorld: Scaling Multi-Agent Video World Model from Single-view Video Data","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31158","citing_title":"Light Interaction: Training-Free Inference Acceleration for Interactive Video World Models","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31734","citing_title":"MemLearner: Learning to Query Context memory for Video World Models","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28816","citing_title":"Gamma-World: Generative Multi-Agent World Modeling Beyond Two Players","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31336","citing_title":"DecMem: Towards Minute-Long Consistent World Generation with Decoupled Memory","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2511.17792","citing_title":"Target-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2512.14614","citing_title":"WorldPlay: Towards Long-Term Geometric Consistency for Real-Time Interactive World Modeling","ref_index":74,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14487","citing_title":"Head Forcing: Long Autoregressive Video Generation via Head Heterogeneity","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08542","citing_title":"Scal3R: Scalable Test-Time Training for Large-Scale 3D Reconstruction","ref_index":94,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06339","citing_title":"Evolution of Video Generative Foundations","ref_index":281,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04514","citing_title":"SuperLocalMemory V3.3: The Living Brain -- Biologically-Inspired Forgetting, Cognitive Quantization, and Multi-Channel Retrieval for Zero-LLM Agent Memory Systems","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14268","citing_title":"HY-World 2.0: A Multi-Modal World Model for Reconstructing, Generating, and Simulating 3D Worlds","ref_index":84,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18564","citing_title":"MultiWorld: Scalable Multi-Agent Multi-View Video World Models","ref_index":62,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF","json":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF.json","graph_json":"https://pith.science/api/pith-number/AQ4KYZHHINCEEMOE7PLBMWJOOF/graph.json","events_json":"https://pith.science/api/pith-number/AQ4KYZHHINCEEMOE7PLBMWJOOF/events.json","paper":"https://pith.science/paper/AQ4KYZHH"},"agent_actions":{"view_html":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF","download_json":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF.json","view_paper":"https://pith.science/paper/AQ4KYZHH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.03141&json=true","fetch_graph":"https://pith.science/api/pith-number/AQ4KYZHHINCEEMOE7PLBMWJOOF/graph.json","fetch_events":"https://pith.science/api/pith-number/AQ4KYZHHINCEEMOE7PLBMWJOOF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF/action/storage_attestation","attest_author":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF/action/author_attestation","sign_citation":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF/action/citation_signature","submit_replication":"https://pith.science/pith/AQ4KYZHHINCEEMOE7PLBMWJOOF/action/replication_record"}},"created_at":"2026-07-05T11:52:20.842648+00:00","updated_at":"2026-07-05T11:52:20.842648+00:00"}