{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ZCMI53QIK5ED7AAUNR4I5S7NH7","short_pith_number":"pith:ZCMI53QI","schema_version":"1.0","canonical_sha256":"c8988eee0857483f80146c788ecbed3fda6e98cc508bfec9bbf32b412081b04f","source":{"kind":"arxiv","id":"2605.30351","version":1},"attestation_state":"computed","paper":{"title":"VideoMLA: Low-Rank Latent KV Cache for Minute-Scale Autoregressive Video Diffusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Adil Kaan Akan, Hidir Yesiltepe, Hoda Eldardiry, Jiazhen Hu, Kaan Oktay, Pinar Yanardag, Tuna Han Salih Meral","submitted_at":"2026-05-28T17:59:57Z","abstract_excerpt":"Long-rollout causal video diffusion has converged on a fixed-size sliding-window KV cache, with recent progress innovating within this layout by changing which tokens occupy the window or how their positions are encoded. The per-head KV layout itself, a dominant contributor to streaming memory and latency, has been mostly left unchanged. In this paper, we present the first study of Multi-Head Latent Attention (MLA) in video diffusion. VideoMLA replaces per-head keys and values with a shared low-rank content latent and a shared decoupled 3D-RoPE positional key, reducing per-token KV memory by 9"},"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":"2605.30351","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-28T17:59:57Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"fbee1eed6dbe724937d9291d169a6d68012468152a5e953058ffb35a993fc488","abstract_canon_sha256":"30292423b4af84ae86232ed0e05dc8ef53095c2c1bd11718d48ebf0649630d30"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T02:06:17.384017Z","signature_b64":"b4Kw7dOwCcGsF3rWUBZQ3zpyLg8sunfOFN73atSEqrbrIXOhbHmnW9X8c+tO9ElTfP2iOAauRkobX9oozBACCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c8988eee0857483f80146c788ecbed3fda6e98cc508bfec9bbf32b412081b04f","last_reissued_at":"2026-05-29T02:06:17.383610Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T02:06:17.383610Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"VideoMLA: Low-Rank Latent KV Cache for Minute-Scale Autoregressive Video Diffusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Adil Kaan Akan, Hidir Yesiltepe, Hoda Eldardiry, Jiazhen Hu, Kaan Oktay, Pinar Yanardag, Tuna Han Salih Meral","submitted_at":"2026-05-28T17:59:57Z","abstract_excerpt":"Long-rollout causal video diffusion has converged on a fixed-size sliding-window KV cache, with recent progress innovating within this layout by changing which tokens occupy the window or how their positions are encoded. The per-head KV layout itself, a dominant contributor to streaming memory and latency, has been mostly left unchanged. In this paper, we present the first study of Multi-Head Latent Attention (MLA) in video diffusion. VideoMLA replaces per-head keys and values with a shared low-rank content latent and a shared decoupled 3D-RoPE positional key, reducing per-token KV memory by 9"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30351","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/2605.30351/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":"2605.30351","created_at":"2026-05-29T02:06:17.383673+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.30351v1","created_at":"2026-05-29T02:06:17.383673+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30351","created_at":"2026-05-29T02:06:17.383673+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZCMI53QIK5ED","created_at":"2026-05-29T02:06:17.383673+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZCMI53QIK5ED7AAU","created_at":"2026-05-29T02:06:17.383673+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZCMI53QI","created_at":"2026-05-29T02:06:17.383673+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/ZCMI53QIK5ED7AAUNR4I5S7NH7","json":"https://pith.science/pith/ZCMI53QIK5ED7AAUNR4I5S7NH7.json","graph_json":"https://pith.science/api/pith-number/ZCMI53QIK5ED7AAUNR4I5S7NH7/graph.json","events_json":"https://pith.science/api/pith-number/ZCMI53QIK5ED7AAUNR4I5S7NH7/events.json","paper":"https://pith.science/paper/ZCMI53QI"},"agent_actions":{"view_html":"https://pith.science/pith/ZCMI53QIK5ED7AAUNR4I5S7NH7","download_json":"https://pith.science/pith/ZCMI53QIK5ED7AAUNR4I5S7NH7.json","view_paper":"https://pith.science/paper/ZCMI53QI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.30351&json=true","fetch_graph":"https://pith.science/api/pith-number/ZCMI53QIK5ED7AAUNR4I5S7NH7/graph.json","fetch_events":"https://pith.science/api/pith-number/ZCMI53QIK5ED7AAUNR4I5S7NH7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZCMI53QIK5ED7AAUNR4I5S7NH7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZCMI53QIK5ED7AAUNR4I5S7NH7/action/storage_attestation","attest_author":"https://pith.science/pith/ZCMI53QIK5ED7AAUNR4I5S7NH7/action/author_attestation","sign_citation":"https://pith.science/pith/ZCMI53QIK5ED7AAUNR4I5S7NH7/action/citation_signature","submit_replication":"https://pith.science/pith/ZCMI53QIK5ED7AAUNR4I5S7NH7/action/replication_record"}},"created_at":"2026-05-29T02:06:17.383673+00:00","updated_at":"2026-05-29T02:06:17.383673+00:00"}