{"paper":{"title":"LIVEditor-14B: Lightning Unified Video Editing via In-Context Sparse Attention","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"In-context Sparse Attention prunes low-saliency context tokens and routes queries by sharpness to cut attention latency by 60 percent while preserving editing quality.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Haopeng Li, Lichen Bai, Shitong Shao, Wenliang Zhong, Yingwei Song, Zeke Xie, Zikai Zhou","submitted_at":"2026-05-06T07:15:29Z","abstract_excerpt":"Video editing has evolved toward In-Context Learning (ICL) paradigms, yet the resulting quadratic attention costs create a critical computational bottleneck. In this work, we propose In-context Sparse Attention (ISA), the first near-lossless empirical sparse framework tailored for ICL video editing. Our design is grounded in two key insights: first, context tokens exhibit significantly lower saliency than source tokens; second, we theoretically prove and empirically validate that Query sharpness correlates with approximation error. Motivated by these findings, ISA implements an efficient pre-s"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ISA is the first near-lossless empirical sparse framework tailored for ICL video editing, achieving ~60% reduction in attention-module latency while surpassing state-of-the-art methods across EditVerseBench, IVE-Bench, and VIE-Bench.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the claimed theoretical proof of query sharpness correlating with approximation error, combined with the empirical pre-selection and dynamic grouping, produces near-lossless results without introducing visual artifacts or requiring per-video retuning on real-world editing tasks.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ISA prunes low-saliency context tokens and routes queries by sharpness to either full or 0-th order Taylor sparse attention, enabling LIVEditor to cut attention latency ~60% while beating prior video editing methods on three benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"In-context Sparse Attention prunes low-saliency context tokens and routes queries by sharpness to cut attention latency by 60 percent while preserving editing quality.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"bd9410f6bd745ec3d1ea455921c4734fd4d7930bd122796505f8f71237c79200"},"source":{"id":"2605.04569","kind":"arxiv","version":2},"verdict":{"id":"a7ab5b81-55eb-4314-b6fa-bc1a3e412596","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T16:32:41.371872Z","strongest_claim":"ISA is the first near-lossless empirical sparse framework tailored for ICL video editing, achieving ~60% reduction in attention-module latency while surpassing state-of-the-art methods across EditVerseBench, IVE-Bench, and VIE-Bench.","one_line_summary":"ISA prunes low-saliency context tokens and routes queries by sharpness to either full or 0-th order Taylor sparse attention, enabling LIVEditor to cut attention latency ~60% while beating prior video editing methods on three benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the claimed theoretical proof of query sharpness correlating with approximation error, combined with the empirical pre-selection and dynamic grouping, produces near-lossless results without introducing visual artifacts or requiring per-video retuning on real-world editing tasks.","pith_extraction_headline":"In-context Sparse Attention prunes low-saliency context tokens and routes queries by sharpness to cut attention latency by 60 percent while preserving editing quality."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.04569/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T11:39:46.501544Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T22:31:20.095333Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T14:20:08.597793Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"6391697ba4302d414bf1e418affe11d3f54dfc8e52726ed8cff251774e6f85c0"},"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"}