{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:ADJSTXNEHHQEQFZIW2RD2VL445","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":"c515fbbf9e9c41cccec4793b0ed0a083e20133de3c46427cba0e8bf131f96a66","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-09-26T17:48:24Z","title_canon_sha256":"638303bcc893c10be0132226122bda9d4b5bc7db58ce39c77753e57801d03740"},"schema_version":"1.0","source":{"id":"2509.22622","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.22622","created_at":"2026-05-17T23:38:53Z"},{"alias_kind":"arxiv_version","alias_value":"2509.22622v2","created_at":"2026-05-17T23:38:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.22622","created_at":"2026-05-17T23:38:53Z"},{"alias_kind":"pith_short_12","alias_value":"ADJSTXNEHHQE","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"ADJSTXNEHHQEQFZI","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"ADJSTXNE","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:fd61739826b618004f6ccb515a1b109391861e7433f8d2a0ffe1f0a640216a49","target":"graph","created_at":"2026-05-17T23:38:53Z","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":"With these key designs, LongLive fine-tunes a 1.3B-parameter short-clip model to minute-long generation in just 32 GPU-days. At inference, LongLive sustains 20.7 FPS on a single NVIDIA H100, achieves strong performance on VBench in both short and long videos. LongLive supports up to 240-second videos on a single H100 GPU."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The assumption that KV-recache combined with short-window attention and frame sink maintains visual consistency and semantic adherence across prompt transitions and long sequences without introducing cumulative artifacts or drift, as this is presented as sufficient based on the described training alignment."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"LongLive is a causal autoregressive video generator that produces up to 240-second interactive videos at 20.7 FPS on one H100 GPU after 32 GPU-days of fine-tuning from a 1.3B short-clip model."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"LongLive turns a short-clip autoregressive model into a real-time system that generates up to 240-second videos at 20.7 FPS while accepting streaming prompt changes."}],"snapshot_sha256":"6da6b3874afa7c3e10e39fc97bb34afa6d505f8e14dc39375d2caecbe5845c07"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"bf6fc593f4075a74190b85611e58d722feecf5b8e6f6537fe49f9a0634ba4ae8"},"paper":{"abstract_excerpt":"We present LongLive, a frame-level autoregressive (AR) framework for real-time and interactive long video generation. Long video generation presents challenges in both efficiency and quality. Diffusion and Diffusion-Forcing models can produce high-quality videos but suffer from low efficiency due to bidirectional attention. Causal attention AR models support KV caching for faster inference, but often degrade in quality on long videos due to memory challenges during long-video training. In addition, beyond static prompt-based generation, interactive capabilities, such as streaming prompt inputs","authors_text":"Enze Xie, Muyang Li, Ruihang Chu, Shuai Yang, Song Han, Wei Huang, Xianbang Wang, Yao Lu, Yicheng Xiao, Yingcong Chen, Yukang Chen, Yuyang Zhao","cross_cats":[],"headline":"LongLive turns a short-clip autoregressive model into a real-time system that generates up to 240-second videos at 20.7 FPS while accepting streaming prompt changes.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-09-26T17:48:24Z","title":"LongLive: Real-time Interactive Long Video Generation"},"references":{"count":108,"internal_anchors":9,"resolved_work":108,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Diffusion forcing: Next-token prediction meets full-sequence diffusion","work_id":"1fb5fc55-61a5-425d-b272-e8246265015a","year":2024},{"cited_arxiv_id":"2504.13074","doi":"","is_internal_anchor":true,"ref_index":2,"title":"SkyReels-V2: Infinite-length Film Generative Model","work_id":"2ce11350-273e-4f0d-ae78-292aa3151060","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Sana-video: Efficient video generation with block linear diffusion transformer","work_id":"37d92c42-ae43-4ef7-adfa-5faf1c48d1a0","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"SEINE: short-to-long video diffusion model for generative transition and prediction","work_id":"fc5b4ca4-5045-4f2e-b2c8-261b55f226f7","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Longlora: Efficient fine-tuning of long-context large language models","work_id":"7f84c8d7-2e59-4005-8191-68c6a956a564","year":2024}],"snapshot_sha256":"939a7220e119ca88778637881b74486a777bfbd7ac7e91910b324b95787baf54"},"source":{"id":"2509.22622","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-15T03:48:41.959401Z","id":"dac721f2-e321-40a9-90f5-04b94f365d3a","model_set":{"reader":"grok-4.3"},"one_line_summary":"LongLive is a causal autoregressive video generator that produces up to 240-second interactive videos at 20.7 FPS on one H100 GPU after 32 GPU-days of fine-tuning from a 1.3B short-clip model.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"LongLive turns a short-clip autoregressive model into a real-time system that generates up to 240-second videos at 20.7 FPS while accepting streaming prompt changes.","strongest_claim":"With these key designs, LongLive fine-tunes a 1.3B-parameter short-clip model to minute-long generation in just 32 GPU-days. At inference, LongLive sustains 20.7 FPS on a single NVIDIA H100, achieves strong performance on VBench in both short and long videos. LongLive supports up to 240-second videos on a single H100 GPU.","weakest_assumption":"The assumption that KV-recache combined with short-window attention and frame sink maintains visual consistency and semantic adherence across prompt transitions and long sequences without introducing cumulative artifacts or drift, as this is presented as sufficient based on the described training alignment."}},"verdict_id":"dac721f2-e321-40a9-90f5-04b94f365d3a"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:a25621e663ddcd995eaac44cf5a0c9ad025a0f5ffdc2f93b92bec08a09034344","target":"record","created_at":"2026-05-17T23:38:53Z","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":"c515fbbf9e9c41cccec4793b0ed0a083e20133de3c46427cba0e8bf131f96a66","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-09-26T17:48:24Z","title_canon_sha256":"638303bcc893c10be0132226122bda9d4b5bc7db58ce39c77753e57801d03740"},"schema_version":"1.0","source":{"id":"2509.22622","kind":"arxiv","version":2}},"canonical_sha256":"00d329dda439e0481728b6a23d557ce7605d4efdcd66dadac49cb2680477fd3f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"00d329dda439e0481728b6a23d557ce7605d4efdcd66dadac49cb2680477fd3f","first_computed_at":"2026-05-17T23:38:53.619263Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:53.619263Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"AOOi4ktb1AHP3M9sUDCfsHbyUG2FdU+qiCBOlyYovZiO+P4RM2MaM9IFRiXzyCdkAMDgJEDBO1nDe7KvCsojAw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:53.619882Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.22622","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a25621e663ddcd995eaac44cf5a0c9ad025a0f5ffdc2f93b92bec08a09034344","sha256:fd61739826b618004f6ccb515a1b109391861e7433f8d2a0ffe1f0a640216a49"],"state_sha256":"9abdc366ca6c393454ff741671a4e8ef637e90377bddf5a8e2d72419ab228b9b"}