{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:LOVGWGVECYQITTLES7U7SADJD3","short_pith_number":"pith:LOVGWGVE","canonical_record":{"source":{"id":"2605.13521","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-13T13:37:45Z","cross_cats_sorted":[],"title_canon_sha256":"b63f54c0e51d9b9a4fba99bdfb61cdd12a3c3c7d88e5c2a9a200b1658cd322f7","abstract_canon_sha256":"0e0dcc0e01adfddb4cbc529f49c8067f6d27be93d7c93868b6cde0f7ff76bde4"},"schema_version":"1.0"},"canonical_sha256":"5baa6b1aa4162089cd6497e9f900691ed0ed048efe296841434fd1ba852547dc","source":{"kind":"arxiv","id":"2605.13521","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13521","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13521v1","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13521","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"pith_short_12","alias_value":"LOVGWGVECYQI","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"LOVGWGVECYQITTLE","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"LOVGWGVE","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:LOVGWGVECYQITTLES7U7SADJD3","target":"record","payload":{"canonical_record":{"source":{"id":"2605.13521","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-13T13:37:45Z","cross_cats_sorted":[],"title_canon_sha256":"b63f54c0e51d9b9a4fba99bdfb61cdd12a3c3c7d88e5c2a9a200b1658cd322f7","abstract_canon_sha256":"0e0dcc0e01adfddb4cbc529f49c8067f6d27be93d7c93868b6cde0f7ff76bde4"},"schema_version":"1.0"},"canonical_sha256":"5baa6b1aa4162089cd6497e9f900691ed0ed048efe296841434fd1ba852547dc","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:24.382359Z","signature_b64":"N31ZwuFLqL9quuaSkQ9nKcsLf0iCdcdTZ2I96PG8j1EDazoNGYanA8Z8vDUWewYGpbuLMYOO+FuWLGVyxwPADQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5baa6b1aa4162089cd6497e9f900691ed0ed048efe296841434fd1ba852547dc","last_reissued_at":"2026-05-18T02:44:24.381883Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:24.381883Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.13521","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"W4pNSsBm82leMLSlhJ0005bEx+wLoHTmMXm6T5gIY5rDzjAKQPuDfGbEc7yxC9m9MZ1bF/XE/1HT1KWdI79uAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T18:14:56.670567Z"},"content_sha256":"c79b7eadd868fe979e6563d5b0e39695669a348a947426b1cb30967d13c3d761","schema_version":"1.0","event_id":"sha256:c79b7eadd868fe979e6563d5b0e39695669a348a947426b1cb30967d13c3d761"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:LOVGWGVECYQITTLES7U7SADJD3","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Granite Embedding Multilingual R2 Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Granite R2 multilingual embedding models achieve state-of-the-art retrieval across more than 200 languages and code.","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Aashka Trivedi, Abraham Daniels, Bhavani Iyer, Jaydeep Sen, Ken Barker, Luis Lastras, Madison Lee, Martin Franz, Meet Doshi, Parul Awasthy, Radu Florian, Riyaz Bhat, Rudra Murthy, Todd Ward, Vignesh P, Vishwajeet Kumar, Yulong Li, Yushu Yang","submitted_at":"2026-05-13T13:37:45Z","abstract_excerpt":"We introduce the multilingual Granite Embedding R2 models, a family of encoder-based embedding models for enterprise-scale dense retrieval across 200+ languages. Extending our English-focused R2 release, these models add enhanced support for 52 languages and programming code, a 32,768-token context window (a 64x expansion over R1), and state-of-the-art overall performance across multilingual and cross-lingual text search, code retrieval, long-document search, and reasoning retrieval datasets. The release consists of two bi-encoder models based on the ModernBERT architecture with an expanded mu"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"state-of-the-art overall performance across multilingual and cross-lingual text search, code retrieval, long-document search, and reasoning retrieval datasets","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The reported benchmarks accurately reflect generalization to real enterprise data without overfitting or benchmark-specific tuning, given that training details and exact evaluation protocols are not provided in the abstract.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Granite Embedding Multilingual R2 releases 311M and 97M parameter bi-encoder models that achieve state-of-the-art retrieval performance on multilingual text, code, long-document, and reasoning datasets.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Granite R2 multilingual embedding models achieve state-of-the-art retrieval across more than 200 languages and code.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a3565c132c73e46e91dc85418edebe357422c5d91b76cc089c42bddc34cdc4f1"},"source":{"id":"2605.13521","kind":"arxiv","version":1},"verdict":{"id":"bc3dc7e3-2a70-432c-b9d9-2366ab9f022c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T18:12:51.598107Z","strongest_claim":"state-of-the-art overall performance across multilingual and cross-lingual text search, code retrieval, long-document search, and reasoning retrieval datasets","one_line_summary":"Granite Embedding Multilingual R2 releases 311M and 97M parameter bi-encoder models that achieve state-of-the-art retrieval performance on multilingual text, code, long-document, and reasoning datasets.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The reported benchmarks accurately reflect generalization to real enterprise data without overfitting or benchmark-specific tuning, given that training details and exact evaluation protocols are not provided in the abstract.","pith_extraction_headline":"Granite R2 multilingual embedding models achieve state-of-the-art retrieval across more than 200 languages and code."},"references":{"count":25,"sample":[{"doi":"","year":null,"title":"Phi-4 Technical Report","work_id":"b6274271-7af9-4ee8-993b-ba1ba4205ba8","ref_index":1,"cited_arxiv_id":"2412.08905","is_internal_anchor":true},{"doi":"","year":2008,"title":"jina-embeddings-v5-text: Task-Targeted Embedding Distillation","work_id":"528e100c-d9f0-4c57-b2f7-b7f63dd5f83e","ref_index":2,"cited_arxiv_id":"2602.15547","is_internal_anchor":true},{"doi":"","year":2008,"title":"3 David Arthur and Sergei Vassilvitskii","work_id":"f59bbab6-3618-4a3e-8d06-bebc6918a16c","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Mmteb: Massive multilingual text embedding benchmark","work_id":"774aa5f1-35ad-4b36-b6cc-5f461cfab347","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.48550/arxiv.2502.13595","year":2021,"title":"Mmteb: Massive multilingual text embedding benchmark","work_id":"774aa5f1-35ad-4b36-b6cc-5f461cfab347","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":25,"snapshot_sha256":"9263ee59bf765105479947748701db83ed8a3ba694c12bf01f48809a0b2a6685","internal_anchors":8},"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"},"verdict_id":"bc3dc7e3-2a70-432c-b9d9-2366ab9f022c"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xMiJKnMu24DWVUeNpcVJvbO8vSAvLsGcuPeBprQYdO517yyVnZE5gg4sY7Ya6Z5D43FJcagTI6zZwwzj/DnqDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T18:14:56.671078Z"},"content_sha256":"a7a0ac5800a62527153900a89948704ecd1366a1d7b3bc341208c15b7cf44e06","schema_version":"1.0","event_id":"sha256:a7a0ac5800a62527153900a89948704ecd1366a1d7b3bc341208c15b7cf44e06"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LOVGWGVECYQITTLES7U7SADJD3/bundle.json","state_url":"https://pith.science/pith/LOVGWGVECYQITTLES7U7SADJD3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LOVGWGVECYQITTLES7U7SADJD3/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-28T18:14:56Z","links":{"resolver":"https://pith.science/pith/LOVGWGVECYQITTLES7U7SADJD3","bundle":"https://pith.science/pith/LOVGWGVECYQITTLES7U7SADJD3/bundle.json","state":"https://pith.science/pith/LOVGWGVECYQITTLES7U7SADJD3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LOVGWGVECYQITTLES7U7SADJD3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:LOVGWGVECYQITTLES7U7SADJD3","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"0e0dcc0e01adfddb4cbc529f49c8067f6d27be93d7c93868b6cde0f7ff76bde4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-13T13:37:45Z","title_canon_sha256":"b63f54c0e51d9b9a4fba99bdfb61cdd12a3c3c7d88e5c2a9a200b1658cd322f7"},"schema_version":"1.0","source":{"id":"2605.13521","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13521","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13521v1","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13521","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"pith_short_12","alias_value":"LOVGWGVECYQI","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"LOVGWGVECYQITTLE","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"LOVGWGVE","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:a7a0ac5800a62527153900a89948704ecd1366a1d7b3bc341208c15b7cf44e06","target":"graph","created_at":"2026-05-18T02:44:24Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"state-of-the-art overall performance across multilingual and cross-lingual text search, code retrieval, long-document search, and reasoning retrieval datasets"},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The reported benchmarks accurately reflect generalization to real enterprise data without overfitting or benchmark-specific tuning, given that training details and exact evaluation protocols are not provided in the abstract."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Granite Embedding Multilingual R2 releases 311M and 97M parameter bi-encoder models that achieve state-of-the-art retrieval performance on multilingual text, code, long-document, and reasoning datasets."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Granite R2 multilingual embedding models achieve state-of-the-art retrieval across more than 200 languages and code."}],"snapshot_sha256":"a3565c132c73e46e91dc85418edebe357422c5d91b76cc089c42bddc34cdc4f1"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"We introduce the multilingual Granite Embedding R2 models, a family of encoder-based embedding models for enterprise-scale dense retrieval across 200+ languages. Extending our English-focused R2 release, these models add enhanced support for 52 languages and programming code, a 32,768-token context window (a 64x expansion over R1), and state-of-the-art overall performance across multilingual and cross-lingual text search, code retrieval, long-document search, and reasoning retrieval datasets. The release consists of two bi-encoder models based on the ModernBERT architecture with an expanded mu","authors_text":"Aashka Trivedi, Abraham Daniels, Bhavani Iyer, Jaydeep Sen, Ken Barker, Luis Lastras, Madison Lee, Martin Franz, Meet Doshi, Parul Awasthy, Radu Florian, Riyaz Bhat, Rudra Murthy, Todd Ward, Vignesh P, Vishwajeet Kumar, Yulong Li, Yushu Yang","cross_cats":[],"headline":"Granite R2 multilingual embedding models achieve state-of-the-art retrieval across more than 200 languages and code.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-13T13:37:45Z","title":"Granite Embedding Multilingual R2 Models"},"references":{"count":25,"internal_anchors":8,"resolved_work":25,"sample":[{"cited_arxiv_id":"2412.08905","doi":"","is_internal_anchor":true,"ref_index":1,"title":"Phi-4 Technical Report","work_id":"b6274271-7af9-4ee8-993b-ba1ba4205ba8","year":null},{"cited_arxiv_id":"2602.15547","doi":"","is_internal_anchor":true,"ref_index":2,"title":"jina-embeddings-v5-text: Task-Targeted Embedding Distillation","work_id":"528e100c-d9f0-4c57-b2f7-b7f63dd5f83e","year":2008},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"3 David Arthur and Sergei Vassilvitskii","work_id":"f59bbab6-3618-4a3e-8d06-bebc6918a16c","year":2008},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Mmteb: Massive multilingual text embedding benchmark","work_id":"774aa5f1-35ad-4b36-b6cc-5f461cfab347","year":null},{"cited_arxiv_id":"","doi":"10.48550/arxiv.2502.13595","is_internal_anchor":false,"ref_index":5,"title":"Mmteb: Massive multilingual text embedding benchmark","work_id":"774aa5f1-35ad-4b36-b6cc-5f461cfab347","year":2021}],"snapshot_sha256":"9263ee59bf765105479947748701db83ed8a3ba694c12bf01f48809a0b2a6685"},"source":{"id":"2605.13521","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T18:12:51.598107Z","id":"bc3dc7e3-2a70-432c-b9d9-2366ab9f022c","model_set":{"reader":"grok-4.3"},"one_line_summary":"Granite Embedding Multilingual R2 releases 311M and 97M parameter bi-encoder models that achieve state-of-the-art retrieval performance on multilingual text, code, long-document, and reasoning datasets.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Granite R2 multilingual embedding models achieve state-of-the-art retrieval across more than 200 languages and code.","strongest_claim":"state-of-the-art overall performance across multilingual and cross-lingual text search, code retrieval, long-document search, and reasoning retrieval datasets","weakest_assumption":"The reported benchmarks accurately reflect generalization to real enterprise data without overfitting or benchmark-specific tuning, given that training details and exact evaluation protocols are not provided in the abstract."}},"verdict_id":"bc3dc7e3-2a70-432c-b9d9-2366ab9f022c"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:c79b7eadd868fe979e6563d5b0e39695669a348a947426b1cb30967d13c3d761","target":"record","created_at":"2026-05-18T02:44:24Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"0e0dcc0e01adfddb4cbc529f49c8067f6d27be93d7c93868b6cde0f7ff76bde4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-05-13T13:37:45Z","title_canon_sha256":"b63f54c0e51d9b9a4fba99bdfb61cdd12a3c3c7d88e5c2a9a200b1658cd322f7"},"schema_version":"1.0","source":{"id":"2605.13521","kind":"arxiv","version":1}},"canonical_sha256":"5baa6b1aa4162089cd6497e9f900691ed0ed048efe296841434fd1ba852547dc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5baa6b1aa4162089cd6497e9f900691ed0ed048efe296841434fd1ba852547dc","first_computed_at":"2026-05-18T02:44:24.381883Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:24.381883Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"N31ZwuFLqL9quuaSkQ9nKcsLf0iCdcdTZ2I96PG8j1EDazoNGYanA8Z8vDUWewYGpbuLMYOO+FuWLGVyxwPADQ==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:24.382359Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13521","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c79b7eadd868fe979e6563d5b0e39695669a348a947426b1cb30967d13c3d761","sha256:a7a0ac5800a62527153900a89948704ecd1366a1d7b3bc341208c15b7cf44e06"],"state_sha256":"1b9e232291c2e236efd423fa9da5388673cde592882caf85989de2b0a45f508a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QQYLQISbaUXjvIFa/v5rWsQPznrEOAYVyCoAMLTi4bdQ+bOtijFCaj27FV1ngOqzI46vfIIX99m+TrtbFmfjBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-28T18:14:56.673454Z","bundle_sha256":"6a81db22531d7d78c7a5825e2391dd2fc60718589d0f0ddce6544638a0493230"}}