{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:LFZ2O5J2FZVUENFA7Q3QJ27GOA","short_pith_number":"pith:LFZ2O5J2","schema_version":"1.0","canonical_sha256":"5973a7753a2e6b4234a0fc3704ebe67007230dc2694fa7a20b9d19be14d9e112","source":{"kind":"arxiv","id":"2208.06366","version":2},"attestation_state":"computed","paper":{"title":"BEiT v2: Masked Image Modeling with Vector-Quantized Visual Tokenizers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Furu Wei, Hangbo Bao, Li Dong, Qixiang Ye, Zhiliang Peng","submitted_at":"2022-08-12T16:48:10Z","abstract_excerpt":"Masked image modeling (MIM) has demonstrated impressive results in self-supervised representation learning by recovering corrupted image patches. However, most existing studies operate on low-level image pixels, which hinders the exploitation of high-level semantics for representation models. In this work, we propose to use a semantic-rich visual tokenizer as the reconstruction target for masked prediction, providing a systematic way to promote MIM from pixel-level to semantic-level. Specifically, we propose vector-quantized knowledge distillation to train the tokenizer, which discretizes a co"},"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":"2208.06366","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-08-12T16:48:10Z","cross_cats_sorted":[],"title_canon_sha256":"716ff212b983ab9a47ade56c57a02144f34dd7c169126f55c7990de1bb60dc5c","abstract_canon_sha256":"8c9061384166c01d26534aadf5ed1614b577a5ec10b9fdb6508d7809243fbca0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:02:44.083724Z","signature_b64":"kcb/4gq4+tCc4Ya9Q3fvcoWqAc8Fx/WZMDP0iEcgueyIXXvJ2TWWUDena3k+9ddnw4QUGCrrZHbFS9qygEtrBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5973a7753a2e6b4234a0fc3704ebe67007230dc2694fa7a20b9d19be14d9e112","last_reissued_at":"2026-07-05T05:02:44.083200Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:02:44.083200Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BEiT v2: Masked Image Modeling with Vector-Quantized Visual Tokenizers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Furu Wei, Hangbo Bao, Li Dong, Qixiang Ye, Zhiliang Peng","submitted_at":"2022-08-12T16:48:10Z","abstract_excerpt":"Masked image modeling (MIM) has demonstrated impressive results in self-supervised representation learning by recovering corrupted image patches. However, most existing studies operate on low-level image pixels, which hinders the exploitation of high-level semantics for representation models. In this work, we propose to use a semantic-rich visual tokenizer as the reconstruction target for masked prediction, providing a systematic way to promote MIM from pixel-level to semantic-level. Specifically, we propose vector-quantized knowledge distillation to train the tokenizer, which discretizes a co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.06366","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/2208.06366/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":"2208.06366","created_at":"2026-07-05T05:02:44.083255+00:00"},{"alias_kind":"arxiv_version","alias_value":"2208.06366v2","created_at":"2026-07-05T05:02:44.083255+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.06366","created_at":"2026-07-05T05:02:44.083255+00:00"},{"alias_kind":"pith_short_12","alias_value":"LFZ2O5J2FZVU","created_at":"2026-07-05T05:02:44.083255+00:00"},{"alias_kind":"pith_short_16","alias_value":"LFZ2O5J2FZVUENFA","created_at":"2026-07-05T05:02:44.083255+00:00"},{"alias_kind":"pith_short_8","alias_value":"LFZ2O5J2","created_at":"2026-07-05T05:02:44.083255+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22077","citing_title":"Morphology-Aware Multimodal Representation Learning for Insect Phylogenetic Reconstruction","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21705","citing_title":"Structural Assessment for Understanding and Guiding Dataset Distillation in Discrete Token Space","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11074","citing_title":"Modeling Complex Behaviors: Multi-Personality Composition and Dynamic Switching in Vision-Language Models","ref_index":88,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31201","citing_title":"ExPLoRe: Expert Patch-Level Loss Routing for Multi-Objective Masked Image Modeling","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27696","citing_title":"Structure over Pixels: Learning Variable-Length Visual Programs","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2506.16950","citing_title":"LAION-C: An Out-of-Distribution Benchmark for Web-Scale Vision Models","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22123","citing_title":"Multilingual Vision-Language Models, A Survey","ref_index":109,"is_internal_anchor":false},{"citing_arxiv_id":"2411.15633","citing_title":"Orthogonal Subspace Decomposition for Generalizable AI-Generated Image Detection","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2410.13848","citing_title":"Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2603.25383","citing_title":"CLIP-RD: Relative Distillation for Efficient CLIP Knowledge Distillation","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2407.07726","citing_title":"PaliGemma: A versatile 3B VLM for transfer","ref_index":106,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07191","citing_title":"Attention Transfer Is Not Universally Effective for Vision Transformers","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA","json":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA.json","graph_json":"https://pith.science/api/pith-number/LFZ2O5J2FZVUENFA7Q3QJ27GOA/graph.json","events_json":"https://pith.science/api/pith-number/LFZ2O5J2FZVUENFA7Q3QJ27GOA/events.json","paper":"https://pith.science/paper/LFZ2O5J2"},"agent_actions":{"view_html":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA","download_json":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA.json","view_paper":"https://pith.science/paper/LFZ2O5J2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2208.06366&json=true","fetch_graph":"https://pith.science/api/pith-number/LFZ2O5J2FZVUENFA7Q3QJ27GOA/graph.json","fetch_events":"https://pith.science/api/pith-number/LFZ2O5J2FZVUENFA7Q3QJ27GOA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA/action/storage_attestation","attest_author":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA/action/author_attestation","sign_citation":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA/action/citation_signature","submit_replication":"https://pith.science/pith/LFZ2O5J2FZVUENFA7Q3QJ27GOA/action/replication_record"}},"created_at":"2026-07-05T05:02:44.083255+00:00","updated_at":"2026-07-05T05:02:44.083255+00:00"}