{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:Z2VNHXRF7UVLN4DQ4IDPN7SUQ2","short_pith_number":"pith:Z2VNHXRF","schema_version":"1.0","canonical_sha256":"ceaad3de25fd2ab6f070e206f6fe5486aa24cf8e55d7f8f54d2a63727ee650f2","source":{"kind":"arxiv","id":"2301.09595","version":2},"attestation_state":"computed","paper":{"title":"Zorro: the masked multimodal transformer","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Adri\\`a Recasens, Andrew Zisserman, Antoine Miech, Drew Jaegle, Jason Lin, Jean-Baptiste Alayrac, Jo\\=ao Carreira, Lucas Smaira, Luyu Wang, Pauline Luc, Ross Hemsley","submitted_at":"2023-01-23T17:51:39Z","abstract_excerpt":"Attention-based models are appealing for multimodal processing because inputs from multiple modalities can be concatenated and fed to a single backbone network - thus requiring very little fusion engineering. The resulting representations are however fully entangled throughout the network, which may not always be desirable: in learning, contrastive audio-visual self-supervised learning requires independent audio and visual features to operate, otherwise learning collapses; in inference, evaluation of audio-visual models should be possible on benchmarks having just audio or just video. In this "},"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":"2301.09595","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-01-23T17:51:39Z","cross_cats_sorted":[],"title_canon_sha256":"aa0bc380c99072a7d9ae3ed286c4c51972925759cf3fbf24ecd9c0c954e86dc3","abstract_canon_sha256":"3e14a0c0625eead350bd3a3d2f23e1ee4c2aa02c971e47524062048abc33c5e3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:44:39.914400Z","signature_b64":"i9xfjldMDqpubPTCfHgZYc+mXWSrsA1kmuG2iXg+EeWuNjIhFAfLduJxpKi794MUyvPphhJYD+zVt66EdSh0Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ceaad3de25fd2ab6f070e206f6fe5486aa24cf8e55d7f8f54d2a63727ee650f2","last_reissued_at":"2026-07-05T05:44:39.913969Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:44:39.913969Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Zorro: the masked multimodal transformer","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Adri\\`a Recasens, Andrew Zisserman, Antoine Miech, Drew Jaegle, Jason Lin, Jean-Baptiste Alayrac, Jo\\=ao Carreira, Lucas Smaira, Luyu Wang, Pauline Luc, Ross Hemsley","submitted_at":"2023-01-23T17:51:39Z","abstract_excerpt":"Attention-based models are appealing for multimodal processing because inputs from multiple modalities can be concatenated and fed to a single backbone network - thus requiring very little fusion engineering. The resulting representations are however fully entangled throughout the network, which may not always be desirable: in learning, contrastive audio-visual self-supervised learning requires independent audio and visual features to operate, otherwise learning collapses; in inference, evaluation of audio-visual models should be possible on benchmarks having just audio or just video. In this "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2301.09595","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/2301.09595/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":"2301.09595","created_at":"2026-07-05T05:44:39.914024+00:00"},{"alias_kind":"arxiv_version","alias_value":"2301.09595v2","created_at":"2026-07-05T05:44:39.914024+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2301.09595","created_at":"2026-07-05T05:44:39.914024+00:00"},{"alias_kind":"pith_short_12","alias_value":"Z2VNHXRF7UVL","created_at":"2026-07-05T05:44:39.914024+00:00"},{"alias_kind":"pith_short_16","alias_value":"Z2VNHXRF7UVLN4DQ","created_at":"2026-07-05T05:44:39.914024+00:00"},{"alias_kind":"pith_short_8","alias_value":"Z2VNHXRF","created_at":"2026-07-05T05:44:39.914024+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.13364","citing_title":"OmniVec2 -- A Novel Transformer based Network for Large Scale Multimodal and Multitask Learning","ref_index":62,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2","json":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2.json","graph_json":"https://pith.science/api/pith-number/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/graph.json","events_json":"https://pith.science/api/pith-number/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/events.json","paper":"https://pith.science/paper/Z2VNHXRF"},"agent_actions":{"view_html":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2","download_json":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2.json","view_paper":"https://pith.science/paper/Z2VNHXRF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2301.09595&json=true","fetch_graph":"https://pith.science/api/pith-number/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/graph.json","fetch_events":"https://pith.science/api/pith-number/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/action/storage_attestation","attest_author":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/action/author_attestation","sign_citation":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/action/citation_signature","submit_replication":"https://pith.science/pith/Z2VNHXRF7UVLN4DQ4IDPN7SUQ2/action/replication_record"}},"created_at":"2026-07-05T05:44:39.914024+00:00","updated_at":"2026-07-05T05:44:39.914024+00:00"}