{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:K4WICMQWKMKOXK7IR4SL5RRRTC","short_pith_number":"pith:K4WICMQW","schema_version":"1.0","canonical_sha256":"572c8132165314ebabe88f24bec6319889e7ec9927c77b48423711b2cdda4738","source":{"kind":"arxiv","id":"2502.06020","version":1},"attestation_state":"computed","paper":{"title":"Temporal Working Memory: Query-Guided Segment Refinement for Enhanced Multimodal Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.MM","cs.SD","eess.AS"],"primary_cat":"cs.CV","authors_text":"Chunhui Zhang, Jiang Gui, Ming Cheng, Peijun Qing, Soroush Vosoughi, Weiyi Wu, Xingjian Diao, Zhongyu Ouyang","submitted_at":"2025-02-09T20:26:30Z","abstract_excerpt":"Multimodal foundation models (MFMs) have demonstrated significant success in tasks such as visual captioning, question answering, and image-text retrieval. However, these models face inherent limitations due to their finite internal capacity, which restricts their ability to process extended temporal sequences, a crucial requirement for comprehensive video and audio analysis. To overcome these challenges, we introduce a specialized cognitive module, temporal working memory (TWM), which aims to enhance the temporal modeling capabilities of MFMs. It selectively retains task-relevant information "},"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":"2502.06020","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-02-09T20:26:30Z","cross_cats_sorted":["cs.MM","cs.SD","eess.AS"],"title_canon_sha256":"3f9a37124008e1905db319bd338cb34d04e2769d42d92f17f134869db35d3d6b","abstract_canon_sha256":"1cecb239f566f9f34a1aaa1e48006cecd30a375d72e02be29a29f5c2dfab6031"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:11:46.551778Z","signature_b64":"0ML5Qk2rf07T8CxyCvRREBEACbaLeNkNItEAp2NQBgkvdLpei0SJ3IokDtnGRi5RxIweWWGVlg9dY4MfjXEQCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"572c8132165314ebabe88f24bec6319889e7ec9927c77b48423711b2cdda4738","last_reissued_at":"2026-07-05T10:11:46.551198Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:11:46.551198Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Temporal Working Memory: Query-Guided Segment Refinement for Enhanced Multimodal Understanding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.MM","cs.SD","eess.AS"],"primary_cat":"cs.CV","authors_text":"Chunhui Zhang, Jiang Gui, Ming Cheng, Peijun Qing, Soroush Vosoughi, Weiyi Wu, Xingjian Diao, Zhongyu Ouyang","submitted_at":"2025-02-09T20:26:30Z","abstract_excerpt":"Multimodal foundation models (MFMs) have demonstrated significant success in tasks such as visual captioning, question answering, and image-text retrieval. However, these models face inherent limitations due to their finite internal capacity, which restricts their ability to process extended temporal sequences, a crucial requirement for comprehensive video and audio analysis. To overcome these challenges, we introduce a specialized cognitive module, temporal working memory (TWM), which aims to enhance the temporal modeling capabilities of MFMs. It selectively retains task-relevant information "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06020","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/2502.06020/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":"2502.06020","created_at":"2026-07-05T10:11:46.551279+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.06020v1","created_at":"2026-07-05T10:11:46.551279+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06020","created_at":"2026-07-05T10:11:46.551279+00:00"},{"alias_kind":"pith_short_12","alias_value":"K4WICMQWKMKO","created_at":"2026-07-05T10:11:46.551279+00:00"},{"alias_kind":"pith_short_16","alias_value":"K4WICMQWKMKOXK7I","created_at":"2026-07-05T10:11:46.551279+00:00"},{"alias_kind":"pith_short_8","alias_value":"K4WICMQW","created_at":"2026-07-05T10:11:46.551279+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2505.20638","citing_title":"Music Audio-Visual Question Answering Requires Specialized Multimodal Designs","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC","json":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC.json","graph_json":"https://pith.science/api/pith-number/K4WICMQWKMKOXK7IR4SL5RRRTC/graph.json","events_json":"https://pith.science/api/pith-number/K4WICMQWKMKOXK7IR4SL5RRRTC/events.json","paper":"https://pith.science/paper/K4WICMQW"},"agent_actions":{"view_html":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC","download_json":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC.json","view_paper":"https://pith.science/paper/K4WICMQW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.06020&json=true","fetch_graph":"https://pith.science/api/pith-number/K4WICMQWKMKOXK7IR4SL5RRRTC/graph.json","fetch_events":"https://pith.science/api/pith-number/K4WICMQWKMKOXK7IR4SL5RRRTC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC/action/storage_attestation","attest_author":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC/action/author_attestation","sign_citation":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC/action/citation_signature","submit_replication":"https://pith.science/pith/K4WICMQWKMKOXK7IR4SL5RRRTC/action/replication_record"}},"created_at":"2026-07-05T10:11:46.551279+00:00","updated_at":"2026-07-05T10:11:46.551279+00:00"}