{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:P4THYW2GJKOKMIVT346FTAT2KS","short_pith_number":"pith:P4THYW2G","schema_version":"1.0","canonical_sha256":"7f267c5b464a9ca622b3df3c59827a54aedc9b05d8ad07bc1bef65c1b5b2d802","source":{"kind":"arxiv","id":"2205.10178","version":2},"attestation_state":"computed","paper":{"title":"Visually-Augmented Language Modeling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Furu Wei, Hao Cheng, Haoyu Song, Jianfeng Gao, Li Dong, Weizhi Wang, XiaoDong Liu, Xifeng Yan","submitted_at":"2022-05-20T13:41:12Z","abstract_excerpt":"Human language is grounded on multimodal knowledge including visual knowledge like colors, sizes, and shapes. However, current large-scale pre-trained language models rely on text-only self-supervised training with massive text data, which precludes them from utilizing relevant visual information when necessary. To address this, we propose a novel pre-training framework, named VaLM, to Visually-augment text tokens with retrieved relevant images for Language Modeling. Specifically, VaLM builds on a novel latent text-image alignment method via an image retrieval module to fetch corresponding ima"},"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":"2205.10178","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-05-20T13:41:12Z","cross_cats_sorted":[],"title_canon_sha256":"8f3cf0239170085bbf780baef02193a6757155e44639a46360d7a01172ed974a","abstract_canon_sha256":"ac5986bf22892dff5c126bcd08501da2ad5a06e712f138814e83888643533c03"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:45:28.543362Z","signature_b64":"T7tLZhQPAL3dc4TLR+KW8mH86T8/NDQTMsfEky2tWmBJPbGbd2pGEq21niWachhwrCAMv7/MVN80/BeaSiJhBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7f267c5b464a9ca622b3df3c59827a54aedc9b05d8ad07bc1bef65c1b5b2d802","last_reissued_at":"2026-07-05T05:45:28.542915Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:45:28.542915Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Visually-Augmented Language Modeling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Furu Wei, Hao Cheng, Haoyu Song, Jianfeng Gao, Li Dong, Weizhi Wang, XiaoDong Liu, Xifeng Yan","submitted_at":"2022-05-20T13:41:12Z","abstract_excerpt":"Human language is grounded on multimodal knowledge including visual knowledge like colors, sizes, and shapes. However, current large-scale pre-trained language models rely on text-only self-supervised training with massive text data, which precludes them from utilizing relevant visual information when necessary. To address this, we propose a novel pre-training framework, named VaLM, to Visually-augment text tokens with retrieved relevant images for Language Modeling. Specifically, VaLM builds on a novel latent text-image alignment method via an image retrieval module to fetch corresponding ima"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.10178","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/2205.10178/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":"2205.10178","created_at":"2026-07-05T05:45:28.542981+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.10178v2","created_at":"2026-07-05T05:45:28.542981+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.10178","created_at":"2026-07-05T05:45:28.542981+00:00"},{"alias_kind":"pith_short_12","alias_value":"P4THYW2GJKOK","created_at":"2026-07-05T05:45:28.542981+00:00"},{"alias_kind":"pith_short_16","alias_value":"P4THYW2GJKOKMIVT","created_at":"2026-07-05T05:45:28.542981+00:00"},{"alias_kind":"pith_short_8","alias_value":"P4THYW2G","created_at":"2026-07-05T05:45:28.542981+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.22933","citing_title":"Augmented Vision-Language Models: A Systematic Review","ref_index":114,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS","json":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS.json","graph_json":"https://pith.science/api/pith-number/P4THYW2GJKOKMIVT346FTAT2KS/graph.json","events_json":"https://pith.science/api/pith-number/P4THYW2GJKOKMIVT346FTAT2KS/events.json","paper":"https://pith.science/paper/P4THYW2G"},"agent_actions":{"view_html":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS","download_json":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS.json","view_paper":"https://pith.science/paper/P4THYW2G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.10178&json=true","fetch_graph":"https://pith.science/api/pith-number/P4THYW2GJKOKMIVT346FTAT2KS/graph.json","fetch_events":"https://pith.science/api/pith-number/P4THYW2GJKOKMIVT346FTAT2KS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS/action/storage_attestation","attest_author":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS/action/author_attestation","sign_citation":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS/action/citation_signature","submit_replication":"https://pith.science/pith/P4THYW2GJKOKMIVT346FTAT2KS/action/replication_record"}},"created_at":"2026-07-05T05:45:28.542981+00:00","updated_at":"2026-07-05T05:45:28.542981+00:00"}