{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:TDGNRRTYBAKZTZ43AAEILPGK73","short_pith_number":"pith:TDGNRRTY","schema_version":"1.0","canonical_sha256":"98ccd8c678081599e79b000885bccafec3828ecd4a3f307121275fbcfbf48d98","source":{"kind":"arxiv","id":"2607.25180","version":1},"attestation_state":"computed","paper":{"title":"Bekko Embedding: Parameter-Efficient Multilingual Retrieval with Ultra-Compact Encoders","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Yuichi Tateno","submitted_at":"2026-07-28T01:13:37Z","abstract_excerpt":"How small can a competitive multilingual retrieval model be? We present Bekko Embedding: its smallest model, bekko-embedding-v1-a8m, has just under 8M Active Parameters (AP) -- the non-embedding parameters that dominate inference compute -- yet on official MMTEB Multilingual v2 Retrieval (nDCG@10) it scores 56.2, above the multilingual-e5 family and BGE-M3 (40x the AP) in our comparison. The higher-quality bekko-embedding-v1-a25m (just under 25M AP) reaches 57.5, on par with gte-multilingual-base, and Multilingual NanoBEIR (14 languages) confirms the trend. Both models handle inputs of up to 8"},"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":"2607.25180","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-07-28T01:13:37Z","cross_cats_sorted":[],"title_canon_sha256":"8d2ba5e37ddd907218fb7ffed6891a7c4cf2cc0903813e1de0cc45fe24e7bc9c","abstract_canon_sha256":"002852700cfa4b8924dae0e5a3232037cd5154f50dbfa8c401b7e35f560dc886"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-29T00:25:07.172211Z","signature_b64":"yEVOq7hccavo2/b++gvuomxTCXNmKxBVbEdbkEqzoJMLRjAODieJ41b1OVTwBWFFnZlMAraSRrr2AE+QcDYUDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"98ccd8c678081599e79b000885bccafec3828ecd4a3f307121275fbcfbf48d98","last_reissued_at":"2026-07-29T00:25:07.171424Z","signature_status":"signed_v1","first_computed_at":"2026-07-29T00:25:07.171424Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bekko Embedding: Parameter-Efficient Multilingual Retrieval with Ultra-Compact Encoders","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Yuichi Tateno","submitted_at":"2026-07-28T01:13:37Z","abstract_excerpt":"How small can a competitive multilingual retrieval model be? We present Bekko Embedding: its smallest model, bekko-embedding-v1-a8m, has just under 8M Active Parameters (AP) -- the non-embedding parameters that dominate inference compute -- yet on official MMTEB Multilingual v2 Retrieval (nDCG@10) it scores 56.2, above the multilingual-e5 family and BGE-M3 (40x the AP) in our comparison. The higher-quality bekko-embedding-v1-a25m (just under 25M AP) reaches 57.5, on par with gte-multilingual-base, and Multilingual NanoBEIR (14 languages) confirms the trend. Both models handle inputs of up to 8"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.25180","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/2607.25180/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":"2607.25180","created_at":"2026-07-29T00:25:07.171828+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.25180v1","created_at":"2026-07-29T00:25:07.171828+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.25180","created_at":"2026-07-29T00:25:07.171828+00:00"},{"alias_kind":"pith_short_12","alias_value":"TDGNRRTYBAKZ","created_at":"2026-07-29T00:25:07.171828+00:00"},{"alias_kind":"pith_short_16","alias_value":"TDGNRRTYBAKZTZ43","created_at":"2026-07-29T00:25:07.171828+00:00"},{"alias_kind":"pith_short_8","alias_value":"TDGNRRTY","created_at":"2026-07-29T00:25:07.171828+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/TDGNRRTYBAKZTZ43AAEILPGK73","json":"https://pith.science/pith/TDGNRRTYBAKZTZ43AAEILPGK73.json","graph_json":"https://pith.science/api/pith-number/TDGNRRTYBAKZTZ43AAEILPGK73/graph.json","events_json":"https://pith.science/api/pith-number/TDGNRRTYBAKZTZ43AAEILPGK73/events.json","paper":"https://pith.science/paper/TDGNRRTY"},"agent_actions":{"view_html":"https://pith.science/pith/TDGNRRTYBAKZTZ43AAEILPGK73","download_json":"https://pith.science/pith/TDGNRRTYBAKZTZ43AAEILPGK73.json","view_paper":"https://pith.science/paper/TDGNRRTY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.25180&json=true","fetch_graph":"https://pith.science/api/pith-number/TDGNRRTYBAKZTZ43AAEILPGK73/graph.json","fetch_events":"https://pith.science/api/pith-number/TDGNRRTYBAKZTZ43AAEILPGK73/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TDGNRRTYBAKZTZ43AAEILPGK73/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TDGNRRTYBAKZTZ43AAEILPGK73/action/storage_attestation","attest_author":"https://pith.science/pith/TDGNRRTYBAKZTZ43AAEILPGK73/action/author_attestation","sign_citation":"https://pith.science/pith/TDGNRRTYBAKZTZ43AAEILPGK73/action/citation_signature","submit_replication":"https://pith.science/pith/TDGNRRTYBAKZTZ43AAEILPGK73/action/replication_record"}},"created_at":"2026-07-29T00:25:07.171828+00:00","updated_at":"2026-07-29T00:25:07.171828+00:00"}