{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UGY6RNCB67KA55ZA3VYYVLBMYW","short_pith_number":"pith:UGY6RNCB","schema_version":"1.0","canonical_sha256":"a1b1e8b441f7d40ef720dd718aac2cc5b82302c63fc3c8ba8a6a93e8e03f1183","source":{"kind":"arxiv","id":"2507.05513","version":1},"attestation_state":"computed","paper":{"title":"Llama Nemoretriever Colembed: Top-Performing Text-Image Retrieval Model","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Benedikt Schifferer, Even Oldridge, Gabriel Moreira, Mengyao Xu, Radek Osmulski, Ronay Ak, Yauhen Babakhin, Zhiding Yu","submitted_at":"2025-07-07T22:20:04Z","abstract_excerpt":"Motivated by the growing demand for retrieval systems that operate across modalities, we introduce llama-nemoretriever-colembed, a unified text-image retrieval model that delivers state-of-the-art performance across multiple benchmarks. We release two model variants, 1B and 3B. The 3B model achieves state of the art performance, scoring NDCG@5 91.0 on ViDoRe V1 and 63.5 on ViDoRe V2, placing first on both leaderboards as of June 27, 2025.\n  Our approach leverages the NVIDIA Eagle2 Vision-Language model (VLM), modifies its architecture by replacing causal attention with bidirectional attention,"},"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":"2507.05513","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2025-07-07T22:20:04Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"34fb3b669afda398e1066accb189ea44419fdef4439c042eb6b7bad893902927","abstract_canon_sha256":"120193e188e461f536c9ced7d0d41aca2eac17419a917cd34149323fc9b03a26"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:33:31.672561Z","signature_b64":"4A3T8j1vPWTItTn/kJowtWceAANBwYvU4wtMl9HBs/l/vt7CsLQRk6A1nrCweptPWk/PcpPLDcark6MPtvapAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a1b1e8b441f7d40ef720dd718aac2cc5b82302c63fc3c8ba8a6a93e8e03f1183","last_reissued_at":"2026-07-05T11:33:31.672129Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:33:31.672129Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Llama Nemoretriever Colembed: Top-Performing Text-Image Retrieval Model","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Benedikt Schifferer, Even Oldridge, Gabriel Moreira, Mengyao Xu, Radek Osmulski, Ronay Ak, Yauhen Babakhin, Zhiding Yu","submitted_at":"2025-07-07T22:20:04Z","abstract_excerpt":"Motivated by the growing demand for retrieval systems that operate across modalities, we introduce llama-nemoretriever-colembed, a unified text-image retrieval model that delivers state-of-the-art performance across multiple benchmarks. We release two model variants, 1B and 3B. The 3B model achieves state of the art performance, scoring NDCG@5 91.0 on ViDoRe V1 and 63.5 on ViDoRe V2, placing first on both leaderboards as of June 27, 2025.\n  Our approach leverages the NVIDIA Eagle2 Vision-Language model (VLM), modifies its architecture by replacing causal attention with bidirectional attention,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.05513","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/2507.05513/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":"2507.05513","created_at":"2026-07-05T11:33:31.672183+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.05513v1","created_at":"2026-07-05T11:33:31.672183+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.05513","created_at":"2026-07-05T11:33:31.672183+00:00"},{"alias_kind":"pith_short_12","alias_value":"UGY6RNCB67KA","created_at":"2026-07-05T11:33:31.672183+00:00"},{"alias_kind":"pith_short_16","alias_value":"UGY6RNCB67KA55ZA","created_at":"2026-07-05T11:33:31.672183+00:00"},{"alias_kind":"pith_short_8","alias_value":"UGY6RNCB","created_at":"2026-07-05T11:33:31.672183+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07654","citing_title":"MM-Matryoshka: Towards Budget-Elastic Visual Document Retrieval via a 2D Multimodal Matryoshka Training Framework","ref_index":91,"is_internal_anchor":false},{"citing_arxiv_id":"2601.21853","citing_title":"LEMUR: Learned Multi-Vector Retrieval","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2509.18095","citing_title":"MetaEmbed: Scaling Multimodal Retrieval at Test-Time with Flexible Late Interaction","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2510.05038","citing_title":"Guided Query Refinement: Multimodal Hybrid Retrieval with Test-Time Optimization","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2511.13415","citing_title":"Attention Grounded Enhancement for Visual Document Retrieval","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2601.21262","citing_title":"CausalEmbed: Auto-Regressive Multi-Vector Generation in Latent Space for Visual Document Embedding","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2601.04720","citing_title":"Qwen3-VL-Embedding and Qwen3-VL-Reranker: A Unified Framework for State-of-the-Art Multimodal Retrieval and Ranking","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08421","citing_title":"Beyond Bag-of-Patches: Learning Global Layout via Textual Supervision for Late-Interaction Visual Document Retrieval","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW","json":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW.json","graph_json":"https://pith.science/api/pith-number/UGY6RNCB67KA55ZA3VYYVLBMYW/graph.json","events_json":"https://pith.science/api/pith-number/UGY6RNCB67KA55ZA3VYYVLBMYW/events.json","paper":"https://pith.science/paper/UGY6RNCB"},"agent_actions":{"view_html":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW","download_json":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW.json","view_paper":"https://pith.science/paper/UGY6RNCB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.05513&json=true","fetch_graph":"https://pith.science/api/pith-number/UGY6RNCB67KA55ZA3VYYVLBMYW/graph.json","fetch_events":"https://pith.science/api/pith-number/UGY6RNCB67KA55ZA3VYYVLBMYW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW/action/storage_attestation","attest_author":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW/action/author_attestation","sign_citation":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW/action/citation_signature","submit_replication":"https://pith.science/pith/UGY6RNCB67KA55ZA3VYYVLBMYW/action/replication_record"}},"created_at":"2026-07-05T11:33:31.672183+00:00","updated_at":"2026-07-05T11:33:31.672183+00:00"}