{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.04581", "kind": "paper", "url": "https://arxiv.org/abs/2607.04581"}, {"id": "gh:tardellirs/mteb-br", "kind": "repo", "url": "https://github.com/tardellirs/mteb-br"}], "authors": ["Tardelli Ronan Coelho Stekel"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.04581", "source_url": "https://arxiv.org/abs/2607.04581"}, "categories": ["General AI", "Information retrieval", "Language & Knowledge", "cs.CL"], "description": {"en": "MTEB-BR is a text embedding benchmark for Brazilian Portuguese with 22 native tasks across seven categories, including classification, clustering, retrieval, and reranking. It evaluates models on native Portuguese data and provides a public leaderboard."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.04581", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.04581", "title": "MTEB-BR"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2607.04581", "modality": null, "name": "MTEB-BR", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "unknown", "origin_record_id": "bm_mteb-br_cc2fe8a1", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "37f009197fee0cbdb675433ba764374999e0c01779751e568f1588d519b0858a", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.04581"}, "publisher": null, "released": "2026-07-06", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.04581", "source_url": "https://arxiv.org/abs/2607.04581"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-04581", "source": "claire_radar", "source_benchmark_id": "2607.04581", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/tardellirs/mteb-br", "githubScope": "benchmark_repo", "githubStars": 12, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 0, "hfPaperUrl": "https://huggingface.co/papers/2607.04581", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "capabilities": ["Information retrieval"], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "1c20f055685d32699e3ebefb3a09fa498ff4594988aebef3557c3af32a27ccb8", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "MTEB-BR is a text embedding benchmark for Brazilian Portuguese with 22 native tasks across seven categories, including classification, clustering, retrieval, and reranking. It evaluates models on native Portuguese data and provides a public leaderboard.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce MTEB-BR, a benchmark of 22 native Brazilian-Portuguese tasks across seven categories (classification, multilabel classification, pair classification, semantic textual similarity, clustering, retrieval, and reranking), admitting only data created or found in Portuguese and excluding translations by construction."}, "familyId": "bmf_16304415cf38", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 2, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/tardellirs/mteb-br", "url": "https://github.com/tardellirs/mteb-br"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/tardellirs/mteb-br", "url": "https://huggingface.co/papers/2607.04581"}], "scope": "benchmark_repo", "url": "https://github.com/tardellirs/mteb-br", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_mteb-br_cc2fe8a1", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/tardellirs/mteb-br", "data": null, "hfPaper": "https://huggingface.co/papers/2607.04581", "paper": "https://arxiv.org/abs/2607.04581", "pdf": "https://arxiv.org/pdf/2607.04581", "project": "https://doi.org/10.5281/zenodo.21087216", "report": "https://arxiv.org/abs/2607.04581"}, "metrics": [], "motivation": "Text embeddings for Portuguese have no dedicated benchmark: evaluation rests on translated corpora such as English MS MARCO or on thin multilingual coverage, with native tasks scattered and unconsolidated.", "name": "MTEB-BR", "oneLine": "MTEB-BR is a text embedding benchmark for Brazilian Portuguese with 22 native tasks across seven categories, including classification, clustering, retrieval, and reranking. It evaluates models on native Portuguese data and provides a public leaderboard.", "paperTitle": "MTEB-BR: A Text Embedding Benchmark for Brazilian Portuguese", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.04581", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-07-06", "source": {"authors": ["Tardelli Ronan Coelho Stekel"], "categories": ["cs.CL", "cs.IR", "cs.LG"], "id": "2607.04581", "title": "MTEB-BR: A Text Embedding Benchmark for Brazilian Portuguese", "type": "arxiv", "url": "https://arxiv.org/abs/2607.04581"}, "sourceUpdatedAt": "2026-07-07T20:54:18Z", "topics": ["cs.CL"], "venueAttempts": [], "whyItMatters": "Portuguese embedding evaluation lacked native benchmarks. This benchmark provides statistically grounded model comparison, showing that multilingual leaderboards only moderately predict Portuguese performance."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
