{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:6QI2M25I7PS73OKHOBI2FRXI2D","short_pith_number":"pith:6QI2M25I","schema_version":"1.0","canonical_sha256":"f411a66ba8fbe5fdb9477051a2c6e8d0c9eec46ddacdc6b93da1513832c60a72","source":{"kind":"arxiv","id":"2503.12496","version":2},"attestation_state":"computed","paper":{"title":"Does Your Vision-Language Model Get Lost in the Long Video Sampling Dilemma?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bei Yu, Bohao Peng, Jiaya Jia, Longxiang Tang, Senqiao Yang, Tianyuan Qu","submitted_at":"2025-03-16T13:12:45Z","abstract_excerpt":"The rise of Large Vision-Language Models (LVLMs) has significantly advanced video understanding. However, efficiently processing long videos remains a challenge due to the ``Sampling Dilemma'': low-density sampling risks missing critical information, while high-density sampling introduces redundancy. To address this issue, we introduce LSDBench, the first benchmark designed to evaluate LVLMs on long-video tasks by constructing high Necessary Sampling Density (NSD) questions, where NSD represents the minimum sampling density required to accurately answer a given question. LSDBench focuses on de"},"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":"2503.12496","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-03-16T13:12:45Z","cross_cats_sorted":[],"title_canon_sha256":"6dabccff7573577e2139e5d846cdf5b99464263b4f916cd77ab83aa2ab5b2c31","abstract_canon_sha256":"163bd4da31f7b7292be0a1768143b2e9a63ca0e36403cbaed9eea86a3e60a655"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:40:39.543628Z","signature_b64":"U7J8LMlW22RUtfgdfvzi4o8G/WRdP2VXNpnwkDapRS7JVP7arQVIPOoxkUxNUvFAc6BiWSPzCOY6olGiQZqqCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f411a66ba8fbe5fdb9477051a2c6e8d0c9eec46ddacdc6b93da1513832c60a72","last_reissued_at":"2026-07-05T10:40:39.543134Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:40:39.543134Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Does Your Vision-Language Model Get Lost in the Long Video Sampling Dilemma?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bei Yu, Bohao Peng, Jiaya Jia, Longxiang Tang, Senqiao Yang, Tianyuan Qu","submitted_at":"2025-03-16T13:12:45Z","abstract_excerpt":"The rise of Large Vision-Language Models (LVLMs) has significantly advanced video understanding. However, efficiently processing long videos remains a challenge due to the ``Sampling Dilemma'': low-density sampling risks missing critical information, while high-density sampling introduces redundancy. To address this issue, we introduce LSDBench, the first benchmark designed to evaluate LVLMs on long-video tasks by constructing high Necessary Sampling Density (NSD) questions, where NSD represents the minimum sampling density required to accurately answer a given question. LSDBench focuses on de"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.12496","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/2503.12496/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":"2503.12496","created_at":"2026-07-05T10:40:39.543191+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.12496v2","created_at":"2026-07-05T10:40:39.543191+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.12496","created_at":"2026-07-05T10:40:39.543191+00:00"},{"alias_kind":"pith_short_12","alias_value":"6QI2M25I7PS7","created_at":"2026-07-05T10:40:39.543191+00:00"},{"alias_kind":"pith_short_16","alias_value":"6QI2M25I7PS73OKH","created_at":"2026-07-05T10:40:39.543191+00:00"},{"alias_kind":"pith_short_8","alias_value":"6QI2M25I","created_at":"2026-07-05T10:40:39.543191+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20726","citing_title":"How Well Can Your Video Model Remember? Measuring Memory-Budget Trade-offs in Long Video Understanding","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14733","citing_title":"Video-Zero: Self-Evolution Video Understanding","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2507.12455","citing_title":"Mitigating Object Hallucinations via Sentence-Level Early Intervention","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03677","citing_title":"Uni-OPD: Unifying On-Policy Distillation with a Dual-Perspective Recipe","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19564","citing_title":"EgoSelf: From Memory to Personalized Egocentric Assistant","ref_index":39,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D","json":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D.json","graph_json":"https://pith.science/api/pith-number/6QI2M25I7PS73OKHOBI2FRXI2D/graph.json","events_json":"https://pith.science/api/pith-number/6QI2M25I7PS73OKHOBI2FRXI2D/events.json","paper":"https://pith.science/paper/6QI2M25I"},"agent_actions":{"view_html":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D","download_json":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D.json","view_paper":"https://pith.science/paper/6QI2M25I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.12496&json=true","fetch_graph":"https://pith.science/api/pith-number/6QI2M25I7PS73OKHOBI2FRXI2D/graph.json","fetch_events":"https://pith.science/api/pith-number/6QI2M25I7PS73OKHOBI2FRXI2D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D/action/storage_attestation","attest_author":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D/action/author_attestation","sign_citation":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D/action/citation_signature","submit_replication":"https://pith.science/pith/6QI2M25I7PS73OKHOBI2FRXI2D/action/replication_record"}},"created_at":"2026-07-05T10:40:39.543191+00:00","updated_at":"2026-07-05T10:40:39.543191+00:00"}