{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:B6L76PTHRTMHIRLF23LF3NANXI","short_pith_number":"pith:B6L76PTH","schema_version":"1.0","canonical_sha256":"0f97ff3e678cd8744565d6d65db40dba3740b611c0d830ce5ba79bd70241d4ea","source":{"kind":"arxiv","id":"2602.16918","version":2},"attestation_state":"computed","paper":{"title":"Xray-Visual Models: Scaling Vision models on Industry Scale Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Aashu Singh, Abhijeet Awasthi, Arkabandhu Chowdhury, Chaitanya Ahuja, Chao Li, David Jacobs, Don Husa, Hao Yuan, Haoyuan Xu, Hongli Xu, Hong-You Chen, Jianpeng Cheng, Jihye Moon, Jun Xiao, Linda Wang, Michael Ge, Michael Hsu, Phong Dingh, Qi Guo, Satya Narayan Shukla, Shlok Mishra, Sreya Dutta Roy, Sumedha Singla, Tsung-Yu Lin, Xiangjun Fan, Yimin Liu, Yonghuan Yang","submitted_at":"2026-02-18T22:22:44Z","abstract_excerpt":"We present Xray-Visual, a unified vision model architecture for large-scale image and video understanding trained on industry-scale social media data. Our model leverages over 15 billion curated image-text pairs and 10 billion video-hashtag pairs from Facebook and Instagram, employing robust data curation pipelines that incorporate balancing and noise suppression strategies to maximize semantic diversity while minimizing label noise. We introduce a three-stage training pipeline that combines self-supervised MAE, semi-supervised hashtag classification, and CLIP-style contrastive learning to joi"},"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":"2602.16918","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-02-18T22:22:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8d0572b7cd8f53e75751bee0d854b773a61c2a950897c7fa5d7b136e4ea56d90","abstract_canon_sha256":"f55b84f8a9e8ed39df0527767569ad7849c1fb159b034b780d8c313c9bb45120"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-15T00:21:16.877013Z","signature_b64":"S2Q0B6uKwy/Iizv+wRJgLoTOQljYOcqnhKrAO7tNMLZlmGJbVMfXC8jYb20UkPG5X43xQnMwoGO7BrE86iF/Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0f97ff3e678cd8744565d6d65db40dba3740b611c0d830ce5ba79bd70241d4ea","last_reissued_at":"2026-07-15T00:21:16.876115Z","signature_status":"signed_v1","first_computed_at":"2026-07-15T00:21:16.876115Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Xray-Visual Models: Scaling Vision models on Industry Scale Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Aashu Singh, Abhijeet Awasthi, Arkabandhu Chowdhury, Chaitanya Ahuja, Chao Li, David Jacobs, Don Husa, Hao Yuan, Haoyuan Xu, Hongli Xu, Hong-You Chen, Jianpeng Cheng, Jihye Moon, Jun Xiao, Linda Wang, Michael Ge, Michael Hsu, Phong Dingh, Qi Guo, Satya Narayan Shukla, Shlok Mishra, Sreya Dutta Roy, Sumedha Singla, Tsung-Yu Lin, Xiangjun Fan, Yimin Liu, Yonghuan Yang","submitted_at":"2026-02-18T22:22:44Z","abstract_excerpt":"We present Xray-Visual, a unified vision model architecture for large-scale image and video understanding trained on industry-scale social media data. Our model leverages over 15 billion curated image-text pairs and 10 billion video-hashtag pairs from Facebook and Instagram, employing robust data curation pipelines that incorporate balancing and noise suppression strategies to maximize semantic diversity while minimizing label noise. We introduce a three-stage training pipeline that combines self-supervised MAE, semi-supervised hashtag classification, and CLIP-style contrastive learning to joi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.16918","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/2602.16918/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":"2602.16918","created_at":"2026-07-15T00:21:16.876547+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.16918v2","created_at":"2026-07-15T00:21:16.876547+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.16918","created_at":"2026-07-15T00:21:16.876547+00:00"},{"alias_kind":"pith_short_12","alias_value":"B6L76PTHRTMH","created_at":"2026-07-15T00:21:16.876547+00:00"},{"alias_kind":"pith_short_16","alias_value":"B6L76PTHRTMHIRLF","created_at":"2026-07-15T00:21:16.876547+00:00"},{"alias_kind":"pith_short_8","alias_value":"B6L76PTH","created_at":"2026-07-15T00:21:16.876547+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/B6L76PTHRTMHIRLF23LF3NANXI","json":"https://pith.science/pith/B6L76PTHRTMHIRLF23LF3NANXI.json","graph_json":"https://pith.science/api/pith-number/B6L76PTHRTMHIRLF23LF3NANXI/graph.json","events_json":"https://pith.science/api/pith-number/B6L76PTHRTMHIRLF23LF3NANXI/events.json","paper":"https://pith.science/paper/B6L76PTH"},"agent_actions":{"view_html":"https://pith.science/pith/B6L76PTHRTMHIRLF23LF3NANXI","download_json":"https://pith.science/pith/B6L76PTHRTMHIRLF23LF3NANXI.json","view_paper":"https://pith.science/paper/B6L76PTH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.16918&json=true","fetch_graph":"https://pith.science/api/pith-number/B6L76PTHRTMHIRLF23LF3NANXI/graph.json","fetch_events":"https://pith.science/api/pith-number/B6L76PTHRTMHIRLF23LF3NANXI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B6L76PTHRTMHIRLF23LF3NANXI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B6L76PTHRTMHIRLF23LF3NANXI/action/storage_attestation","attest_author":"https://pith.science/pith/B6L76PTHRTMHIRLF23LF3NANXI/action/author_attestation","sign_citation":"https://pith.science/pith/B6L76PTHRTMHIRLF23LF3NANXI/action/citation_signature","submit_replication":"https://pith.science/pith/B6L76PTHRTMHIRLF23LF3NANXI/action/replication_record"}},"created_at":"2026-07-15T00:21:16.876547+00:00","updated_at":"2026-07-15T00:21:16.876547+00:00"}