{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UXFXZH2DU5QVLZ4AEWRRDPVGLS","short_pith_number":"pith:UXFXZH2D","schema_version":"1.0","canonical_sha256":"a5cb7c9f43a76155e78025a311bea65cbb05285501c8a6ca3eacf9a8ab3056f5","source":{"kind":"arxiv","id":"2502.21271","version":1},"attestation_state":"computed","paper":{"title":"Adaptive Keyframe Sampling for Long Video Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Jianbin Jiao, Jihao Qiu, Lingxi Xie, Qixiang Ye, Xi Tang, Yunjie Tian","submitted_at":"2025-02-28T17:46:29Z","abstract_excerpt":"Multimodal large language models (MLLMs) have enabled open-world visual understanding by injecting visual input as extra tokens into large language models (LLMs) as contexts. However, when the visual input changes from a single image to a long video, the above paradigm encounters difficulty because the vast amount of video tokens has significantly exceeded the maximal capacity of MLLMs. Therefore, existing video-based MLLMs are mostly established upon sampling a small portion of tokens from input data, which can cause key information to be lost and thus produce incorrect answers. This paper pr"},"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":"2502.21271","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-02-28T17:46:29Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"94e3e54a97fd884676bb0432345322bc994305fb7a0ff2cd2bb6cb1825f32e74","abstract_canon_sha256":"629c10222f7057a5c4bbf7695f0a9a0878937ba4038c592fa98d66030d5d36de"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:21:47.405741Z","signature_b64":"HJ3jYsegr/q9iSXy3bfaBEezh7pO3b2Hr+QSsmYqe0j8zqxTA57q9Ep0aOESvZhkUbC/TTXQs0YrN2JTgagkDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a5cb7c9f43a76155e78025a311bea65cbb05285501c8a6ca3eacf9a8ab3056f5","last_reissued_at":"2026-07-05T10:21:47.405221Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:21:47.405221Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adaptive Keyframe Sampling for Long Video Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Jianbin Jiao, Jihao Qiu, Lingxi Xie, Qixiang Ye, Xi Tang, Yunjie Tian","submitted_at":"2025-02-28T17:46:29Z","abstract_excerpt":"Multimodal large language models (MLLMs) have enabled open-world visual understanding by injecting visual input as extra tokens into large language models (LLMs) as contexts. However, when the visual input changes from a single image to a long video, the above paradigm encounters difficulty because the vast amount of video tokens has significantly exceeded the maximal capacity of MLLMs. Therefore, existing video-based MLLMs are mostly established upon sampling a small portion of tokens from input data, which can cause key information to be lost and thus produce incorrect answers. This paper pr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.21271","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/2502.21271/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":"2502.21271","created_at":"2026-07-05T10:21:47.405283+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.21271v1","created_at":"2026-07-05T10:21:47.405283+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.21271","created_at":"2026-07-05T10:21:47.405283+00:00"},{"alias_kind":"pith_short_12","alias_value":"UXFXZH2DU5QV","created_at":"2026-07-05T10:21:47.405283+00:00"},{"alias_kind":"pith_short_16","alias_value":"UXFXZH2DU5QVLZ4A","created_at":"2026-07-05T10:21:47.405283+00:00"},{"alias_kind":"pith_short_8","alias_value":"UXFXZH2D","created_at":"2026-07-05T10:21:47.405283+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.31029","citing_title":"PEEK: Picking Essential frames via Efficient Knowledge distillation","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16740","citing_title":"TRACE: Evidence Grounding-Guided Multi-Video Event Understanding and Claim Generation","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29472","citing_title":"Agent-Computer Observation Interfaces Enable Dynamic Computer Use","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27318","citing_title":"Q-GeoMem: Question-Guided Geometric Memory for Video Spatial Reasoning","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2505.15269","citing_title":"LiveVLM: Efficient Online Video Understanding via Streaming-Oriented KV Cache and Retrieval","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16740","citing_title":"TRACE: Evidence Grounding-Guided Multi-Video Event Understanding and Claim Generation","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2512.08410","citing_title":"Towards Effective Long Video Understanding of Multimodal Large Language Models via One-shot Clip Retrieval","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2601.14594","citing_title":"LFS: Learnable Frame Selector for Event-Aware and Temporally Diverse Video Captioning","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09223","citing_title":"CREST: Curvature-Regulated Event-Centric Sampling for Efficient Long-Video Understanding","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS","json":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS.json","graph_json":"https://pith.science/api/pith-number/UXFXZH2DU5QVLZ4AEWRRDPVGLS/graph.json","events_json":"https://pith.science/api/pith-number/UXFXZH2DU5QVLZ4AEWRRDPVGLS/events.json","paper":"https://pith.science/paper/UXFXZH2D"},"agent_actions":{"view_html":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS","download_json":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS.json","view_paper":"https://pith.science/paper/UXFXZH2D","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.21271&json=true","fetch_graph":"https://pith.science/api/pith-number/UXFXZH2DU5QVLZ4AEWRRDPVGLS/graph.json","fetch_events":"https://pith.science/api/pith-number/UXFXZH2DU5QVLZ4AEWRRDPVGLS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS/action/storage_attestation","attest_author":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS/action/author_attestation","sign_citation":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS/action/citation_signature","submit_replication":"https://pith.science/pith/UXFXZH2DU5QVLZ4AEWRRDPVGLS/action/replication_record"}},"created_at":"2026-07-05T10:21:47.405283+00:00","updated_at":"2026-07-05T10:21:47.405283+00:00"}