{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:KIAWPVMOBRGZMN3B4BYSHEZIYJ","short_pith_number":"pith:KIAWPVMO","canonical_record":{"source":{"id":"2601.23286","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-30T18:59:57Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"007ad995694ff2255cd9995d4b9a9e0c5a0796435b0a63eb364eba611fbe179c","abstract_canon_sha256":"6cd1add77c7d5ca75e04b8bff6d1b3827e81ad8e70642d24800cbd724e202650"},"schema_version":"1.0"},"canonical_sha256":"520167d58e0c4d963761e071239328c242bb6c1af6d3d12151c1cb28c1613c57","source":{"kind":"arxiv","id":"2601.23286","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.23286","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"arxiv_version","alias_value":"2601.23286v4","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.23286","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"pith_short_12","alias_value":"KIAWPVMOBRGZ","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"pith_short_16","alias_value":"KIAWPVMOBRGZMN3B","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"pith_short_8","alias_value":"KIAWPVMO","created_at":"2026-06-09T02:07:18Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:KIAWPVMOBRGZMN3B4BYSHEZIYJ","target":"record","payload":{"canonical_record":{"source":{"id":"2601.23286","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-30T18:59:57Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"007ad995694ff2255cd9995d4b9a9e0c5a0796435b0a63eb364eba611fbe179c","abstract_canon_sha256":"6cd1add77c7d5ca75e04b8bff6d1b3827e81ad8e70642d24800cbd724e202650"},"schema_version":"1.0"},"canonical_sha256":"520167d58e0c4d963761e071239328c242bb6c1af6d3d12151c1cb28c1613c57","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T02:07:18.733201Z","signature_b64":"HBXZiqviHsVGJb2k013ZtDnLlioeSGfG0/C7KZXSz6HlUFWJOaYC/QdwNGOmIF9j886IOco0l5syi6hF0aYRCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"520167d58e0c4d963761e071239328c242bb6c1af6d3d12151c1cb28c1613c57","last_reissued_at":"2026-06-09T02:07:18.732244Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T02:07:18.732244Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2601.23286","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-09T02:07:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zBLHDAG6WJkUyXjgUdlX0LyUWWtYJYdU+lMd2hjyGm5tesm3HLBj/CIeI66YgOfx3icGChgD93mx6DVzhwaNDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T00:44:28.898837Z"},"content_sha256":"f8820334915a51b3f13a5094bdcb9b31498e67f25839cbee81d35cda5c1fa9f5","schema_version":"1.0","event_id":"sha256:f8820334915a51b3f13a5094bdcb9b31498e67f25839cbee81d35cda5c1fa9f5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:KIAWPVMOBRGZMN3B4BYSHEZIYJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"VideoGPA: Distilling Geometry Priors for 3D-Consistent Video Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Video diffusion models gain 3D consistency by distilling preference signals from a geometry foundation model.","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Aman Agarwal, Hongyang Du, Jingcheng Ni, Junjie Ye, Randall Balestriero, Runhao Li, Xiaoyan Cong, Yue Wang, Zekun Li, Zeqi Zhou","submitted_at":"2026-01-30T18:59:57Z","abstract_excerpt":"While recent video diffusion models (VDMs) produce visually impressive results, they fundamentally struggle to maintain 3D structural consistency, often resulting in object deformation or spatial drift. We hypothesize that these failures arise because standard denoising objectives lack explicit incentives for geometric coherence. To address this, we introduce VideoGPA (Video Geometric Preference Alignment), a data-efficient self-supervised framework that leverages a geometry foundation model to automatically derive dense preference signals that guide VDMs via Direct Preference Optimization (DP"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"VideoGPA significantly enhances temporal stability, geometric plausibility, and motion coherence using minimal preference pairs, consistently outperforming state-of-the-art baselines in extensive experiments.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That a geometry foundation model can reliably derive accurate dense preference signals for 3D consistency that effectively guide the video diffusion model without introducing its own errors or biases.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"VideoGPA distills geometry priors via self-supervised DPO to enhance 3D consistency, temporal stability, and motion coherence in video diffusion models.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Video diffusion models gain 3D consistency by distilling preference signals from a geometry foundation model.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"c026541c6e3d958e3d52ed949e7569714367cc3bfdf750b26288557ab40d49d3"},"source":{"id":"2601.23286","kind":"arxiv","version":4},"verdict":{"id":"1a7b954a-062f-43c9-9493-a63939955478","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T09:15:06.724264Z","strongest_claim":"VideoGPA significantly enhances temporal stability, geometric plausibility, and motion coherence using minimal preference pairs, consistently outperforming state-of-the-art baselines in extensive experiments.","one_line_summary":"VideoGPA distills geometry priors via self-supervised DPO to enhance 3D consistency, temporal stability, and motion coherence in video diffusion models.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That a geometry foundation model can reliably derive accurate dense preference signals for 3D consistency that effectively guide the video diffusion model without introducing its own errors or biases.","pith_extraction_headline":"Video diffusion models gain 3D consistency by distilling preference signals from a geometry foundation model."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2601.23286/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":2,"snapshot_sha256":"0214bd2eff3743bac3a53a25d94bb6038f02ea1d3ee6d0de05a8205e0ea5108f"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"1a7b954a-062f-43c9-9493-a63939955478"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-09T02:07:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"f94q8wVBE2nJdyxrbu8+LxK5Lk/APcxLjPo8N6pIkJG5vfj7pS7BCYVdLL6OjUsMe9k1LqolhWJXGxpbIql9BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T00:44:28.899801Z"},"content_sha256":"e5f55142ab7066341913644cd881077c9591e3ad2eaebfa8f75bfdaca562b281","schema_version":"1.0","event_id":"sha256:e5f55142ab7066341913644cd881077c9591e3ad2eaebfa8f75bfdaca562b281"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/KIAWPVMOBRGZMN3B4BYSHEZIYJ/bundle.json","state_url":"https://pith.science/pith/KIAWPVMOBRGZMN3B4BYSHEZIYJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/KIAWPVMOBRGZMN3B4BYSHEZIYJ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-06T00:44:28Z","links":{"resolver":"https://pith.science/pith/KIAWPVMOBRGZMN3B4BYSHEZIYJ","bundle":"https://pith.science/pith/KIAWPVMOBRGZMN3B4BYSHEZIYJ/bundle.json","state":"https://pith.science/pith/KIAWPVMOBRGZMN3B4BYSHEZIYJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/KIAWPVMOBRGZMN3B4BYSHEZIYJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:KIAWPVMOBRGZMN3B4BYSHEZIYJ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"6cd1add77c7d5ca75e04b8bff6d1b3827e81ad8e70642d24800cbd724e202650","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-30T18:59:57Z","title_canon_sha256":"007ad995694ff2255cd9995d4b9a9e0c5a0796435b0a63eb364eba611fbe179c"},"schema_version":"1.0","source":{"id":"2601.23286","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2601.23286","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"arxiv_version","alias_value":"2601.23286v4","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.23286","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"pith_short_12","alias_value":"KIAWPVMOBRGZ","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"pith_short_16","alias_value":"KIAWPVMOBRGZMN3B","created_at":"2026-06-09T02:07:18Z"},{"alias_kind":"pith_short_8","alias_value":"KIAWPVMO","created_at":"2026-06-09T02:07:18Z"}],"graph_snapshots":[{"event_id":"sha256:e5f55142ab7066341913644cd881077c9591e3ad2eaebfa8f75bfdaca562b281","target":"graph","created_at":"2026-06-09T02:07:18Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"VideoGPA significantly enhances temporal stability, geometric plausibility, and motion coherence using minimal preference pairs, consistently outperforming state-of-the-art baselines in extensive experiments."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That a geometry foundation model can reliably derive accurate dense preference signals for 3D consistency that effectively guide the video diffusion model without introducing its own errors or biases."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"VideoGPA distills geometry priors via self-supervised DPO to enhance 3D consistency, temporal stability, and motion coherence in video diffusion models."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Video diffusion models gain 3D consistency by distilling preference signals from a geometry foundation model."}],"snapshot_sha256":"c026541c6e3d958e3d52ed949e7569714367cc3bfdf750b26288557ab40d49d3"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"0214bd2eff3743bac3a53a25d94bb6038f02ea1d3ee6d0de05a8205e0ea5108f"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2601.23286/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"While recent video diffusion models (VDMs) produce visually impressive results, they fundamentally struggle to maintain 3D structural consistency, often resulting in object deformation or spatial drift. We hypothesize that these failures arise because standard denoising objectives lack explicit incentives for geometric coherence. To address this, we introduce VideoGPA (Video Geometric Preference Alignment), a data-efficient self-supervised framework that leverages a geometry foundation model to automatically derive dense preference signals that guide VDMs via Direct Preference Optimization (DP","authors_text":"Aman Agarwal, Hongyang Du, Jingcheng Ni, Junjie Ye, Randall Balestriero, Runhao Li, Xiaoyan Cong, Yue Wang, Zekun Li, Zeqi Zhou","cross_cats":["cs.AI","cs.LG"],"headline":"Video diffusion models gain 3D consistency by distilling preference signals from a geometry foundation model.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-30T18:59:57Z","title":"VideoGPA: Distilling Geometry Priors for 3D-Consistent Video Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.23286","kind":"arxiv","version":4},"verdict":{"created_at":"2026-05-16T09:15:06.724264Z","id":"1a7b954a-062f-43c9-9493-a63939955478","model_set":{"reader":"grok-4.3"},"one_line_summary":"VideoGPA distills geometry priors via self-supervised DPO to enhance 3D consistency, temporal stability, and motion coherence in video diffusion models.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Video diffusion models gain 3D consistency by distilling preference signals from a geometry foundation model.","strongest_claim":"VideoGPA significantly enhances temporal stability, geometric plausibility, and motion coherence using minimal preference pairs, consistently outperforming state-of-the-art baselines in extensive experiments.","weakest_assumption":"That a geometry foundation model can reliably derive accurate dense preference signals for 3D consistency that effectively guide the video diffusion model without introducing its own errors or biases."}},"verdict_id":"1a7b954a-062f-43c9-9493-a63939955478"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:f8820334915a51b3f13a5094bdcb9b31498e67f25839cbee81d35cda5c1fa9f5","target":"record","created_at":"2026-06-09T02:07:18Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"6cd1add77c7d5ca75e04b8bff6d1b3827e81ad8e70642d24800cbd724e202650","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-01-30T18:59:57Z","title_canon_sha256":"007ad995694ff2255cd9995d4b9a9e0c5a0796435b0a63eb364eba611fbe179c"},"schema_version":"1.0","source":{"id":"2601.23286","kind":"arxiv","version":4}},"canonical_sha256":"520167d58e0c4d963761e071239328c242bb6c1af6d3d12151c1cb28c1613c57","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"520167d58e0c4d963761e071239328c242bb6c1af6d3d12151c1cb28c1613c57","first_computed_at":"2026-06-09T02:07:18.732244Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-09T02:07:18.732244Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"HBXZiqviHsVGJb2k013ZtDnLlioeSGfG0/C7KZXSz6HlUFWJOaYC/QdwNGOmIF9j886IOco0l5syi6hF0aYRCQ==","signature_status":"signed_v1","signed_at":"2026-06-09T02:07:18.733201Z","signed_message":"canonical_sha256_bytes"},"source_id":"2601.23286","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f8820334915a51b3f13a5094bdcb9b31498e67f25839cbee81d35cda5c1fa9f5","sha256:e5f55142ab7066341913644cd881077c9591e3ad2eaebfa8f75bfdaca562b281"],"state_sha256":"d9c5af8f590e38fc8c6e4332e86ecb4b5d02aa5179e65111bdb751f072a9a000"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eYtR3SPkoW+3T9dXr7jS3AeoKg9uX0AgIOanRsqlMLKvfbbLZ0EtJ67qccRgH+rX3GjYXH8cr8bhfdJ5TBZUAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T00:44:28.906482Z","bundle_sha256":"786d9e3b8a1e96649e31bf0f71077327eb4cf7d329ccf89e25d8efeff724bfd6"}}