{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:BLDT2QBLPTADLSEJGXC7HACBNN","short_pith_number":"pith:BLDT2QBL","schema_version":"1.0","canonical_sha256":"0ac73d402b7cc035c88935c5f380416b65af785d972799d95c2dfcc28914b41d","source":{"kind":"arxiv","id":"2112.07074","version":1},"attestation_state":"computed","paper":{"title":"Towards a Unified Foundation Model: Jointly Pre-Training Transformers on Unpaired Images and Text","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Boqing Gong, Dan Kondratyuk, Matthew Brown, Ming-Hsuan Yang, Qing Li, Xianzhi Du, Yin Cui","submitted_at":"2021-12-14T00:20:55Z","abstract_excerpt":"In this paper, we explore the possibility of building a unified foundation model that can be adapted to both vision-only and text-only tasks. Starting from BERT and ViT, we design a unified transformer consisting of modality-specific tokenizers, a shared transformer encoder, and task-specific output heads. To efficiently pre-train the proposed model jointly on unpaired images and text, we propose two novel techniques: (i) We employ the separately-trained BERT and ViT models as teachers and apply knowledge distillation to provide additional, accurate supervision signals for the joint training; "},"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":"2112.07074","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-12-14T00:20:55Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"9058cc91ef1d319629353b97047175007ddc2d02115df2427f3840c6899e5be4","abstract_canon_sha256":"63bd291f7586ac522c336e3c99de9b7fde7a1ef7b8280b64904068991dcf8721"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:40:20.608934Z","signature_b64":"A6I2Y1NF9q4T3JBJ1A9E4SYbiOy0PkXZL9QoTlsUCxtwi+XnZj/Lfz0yWOYBsxtDvxsaMilzupCFvXfVa/AcCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0ac73d402b7cc035c88935c5f380416b65af785d972799d95c2dfcc28914b41d","last_reissued_at":"2026-07-05T03:40:20.608522Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:40:20.608522Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards a Unified Foundation Model: Jointly Pre-Training Transformers on Unpaired Images and Text","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Boqing Gong, Dan Kondratyuk, Matthew Brown, Ming-Hsuan Yang, Qing Li, Xianzhi Du, Yin Cui","submitted_at":"2021-12-14T00:20:55Z","abstract_excerpt":"In this paper, we explore the possibility of building a unified foundation model that can be adapted to both vision-only and text-only tasks. Starting from BERT and ViT, we design a unified transformer consisting of modality-specific tokenizers, a shared transformer encoder, and task-specific output heads. To efficiently pre-train the proposed model jointly on unpaired images and text, we propose two novel techniques: (i) We employ the separately-trained BERT and ViT models as teachers and apply knowledge distillation to provide additional, accurate supervision signals for the joint training; "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.07074","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/2112.07074/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":"2112.07074","created_at":"2026-07-05T03:40:20.608582+00:00"},{"alias_kind":"arxiv_version","alias_value":"2112.07074v1","created_at":"2026-07-05T03:40:20.608582+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.07074","created_at":"2026-07-05T03:40:20.608582+00:00"},{"alias_kind":"pith_short_12","alias_value":"BLDT2QBLPTAD","created_at":"2026-07-05T03:40:20.608582+00:00"},{"alias_kind":"pith_short_16","alias_value":"BLDT2QBLPTADLSEJ","created_at":"2026-07-05T03:40:20.608582+00:00"},{"alias_kind":"pith_short_8","alias_value":"BLDT2QBL","created_at":"2026-07-05T03:40:20.608582+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN","json":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN.json","graph_json":"https://pith.science/api/pith-number/BLDT2QBLPTADLSEJGXC7HACBNN/graph.json","events_json":"https://pith.science/api/pith-number/BLDT2QBLPTADLSEJGXC7HACBNN/events.json","paper":"https://pith.science/paper/BLDT2QBL"},"agent_actions":{"view_html":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN","download_json":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN.json","view_paper":"https://pith.science/paper/BLDT2QBL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2112.07074&json=true","fetch_graph":"https://pith.science/api/pith-number/BLDT2QBLPTADLSEJGXC7HACBNN/graph.json","fetch_events":"https://pith.science/api/pith-number/BLDT2QBLPTADLSEJGXC7HACBNN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN/action/storage_attestation","attest_author":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN/action/author_attestation","sign_citation":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN/action/citation_signature","submit_replication":"https://pith.science/pith/BLDT2QBLPTADLSEJGXC7HACBNN/action/replication_record"}},"created_at":"2026-07-05T03:40:20.608582+00:00","updated_at":"2026-07-05T03:40:20.608582+00:00"}