{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.02379", "kind": "paper", "url": "https://arxiv.org/abs/2609.02379"}, {"id": "gh:grecomt/multighostbench", "kind": "repo", "url": "https://github.com/GrecoMT/MultiGhostBench"}], "authors": ["Matteo Greco", "Anudeex Shetty", "Andrea Tagarelli", "Jey Han Lau"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.02379", "source_url": "https://arxiv.org/abs/2609.02379"}, "categories": ["General AI", "Language & Knowledge", "cs.CL"], "description": {"en": "Contains 928 machine-generated books across six languages and three scripts for evaluating long-form LLM-generated text attribution under distribution shifts."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.02379", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.02379", "title": "MultiGhostBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["The University of Melbourne", "University of Calabria"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2609.02379", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2609.02379", "source_url": "https://openalex.org/W7207603047"}, "key": "claire-radar:2609.02379", "modality": null, "name": "MultiGhostBench", "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": "false", "origin_record_id": "bm_multighostbench_4adb4861", "origin_source": "arxiv", "recognition_confidence": "0.95", "record_sha256": "5e52a6baa42ad6ff3f3c24c2d25224ac6fb3a77282cf21ca05b813418c40e942", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-04T00:46:59.007339Z", "source_url": "https://arxiv.org/abs/2609.02379"}, "publisher": null, "released": "2026-09-02", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.02379", "source_url": "https://arxiv.org/abs/2609.02379"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-02379", "source": "claire_radar", "source_benchmark_id": "2609.02379", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-09-03", "attemptedAt": "2026-09-03T01:40:48.194060Z", "githubRepo": "https://github.com/GrecoMT/MultiGhostBench", "githubScope": "benchmark_repo", "githubStars": 1, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.02379", "observedAt": "2026-09-03T01:40:48.194060Z", "signalStatus": {"githubStars": {"observedAt": "2026-09-03T01:40:48.194060Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-03T01:40:48.194060Z", "state": "unavailable"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Multilingual long-form attribution is a growing subfield, but the niche benchmark may attract moderate attention among text detection researchers.", "score": 35}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-04T00:46:59.007339Z", "inputHash": "650c473f531078f8ea1164d88c567aa8425e3aa939d8c8abed4488a83e41b70c", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "MultiGhostBench: A Multilingual Benchmark for Long-Form LLM-Generated Text Attribution under Distribution Shifts", "canonicalNameSource": "paper_title", "decisionReason": "The benchmark is explicitly named and the GitHub repository is available with code and dataset references, offering a public reusable path, though scoring details are not provided.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-04T00:46:59.007339Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Contains 928 machine-generated books across six languages and three scripts for evaluating long-form LLM-generated text attribution under distribution shifts.", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "public artifact URL"], "snippet": "We introduce MultiGhostBench, a multilingual benchmark comprising 928 books generated by five recent LLMs across six languages and three scripts, with an average length of approximately 59K words per book."}, "familyId": "bmf_424de7215466", "firstSeenAt": "2026-09-03", "heat": null, "id": "bm_multighostbench_4adb4861", "indexedAt": "2026-09-03T00:55:28.123857Z", "industrySectors": [], "links": {"code": "https://github.com/GrecoMT/MultiGhostBench", "data": null, "paper": "https://arxiv.org/abs/2609.02379", "pdf": "https://arxiv.org/pdf/2609.02379", "project": null, "report": "https://arxiv.org/abs/2609.02379"}, "metrics": [], "motivation": "While existing work on LLM authorship attribution (AA) has made progress, available benchmarks remain limited, often focusing on English, controlled settings, or relatively outdated models, with the few multilingual studies considering only relatively short texts.", "name": "MultiGhostBench", "oneLine": "Contains 928 machine-generated books across six languages and three scripts for evaluating long-form LLM-generated text attribution under distribution shifts.", "paperTitle": "MultiGhostBench: A Multilingual Benchmark for Long-Form LLM-Generated Text Attribution under Distribution Shifts", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.02379", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-03T00:55:28.123857Z"}, "publications": [], "publishers": [{"name": "GrecoMT", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/GrecoMT/MultiGhostBench"}], "ranking": {"30d": {"confidence": "Low", "coverage": 0.55, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": 0.10323645750017778, "value": 1}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 93, "score": 28}, "method": "attention-ranking-v13", "rank": 93, "score": 28}, "90d": {"confidence": "Low", "coverage": 0.55, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": 0.09423341498807918, "value": 1}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 254, "score": 28}, "method": "attention-ranking-v13", "rank": 254, "score": 28}, "today": {"confidence": "Low", "coverage": 0.25, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": 1, "historyDate": "2026-09-02", "rank": null, "score": null, "windowDays": 1}, "level": {"components": {"githubStars": {"normalization": "percentile", "normalized": 0.5, "percentile": 0.5, "value": 1}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "llmAttentionForecast": {"maxContribution": 0.25, "percentile": 0.58, "role": "bonus", "value": 58}}, "confidence": "Low", "coverage": 0.25, "rank": 2, "score": 52}, "method": "attention-ranking-v13", "rank": 2, "score": 52}}, "readiness": "Runnable", "recognitionConfidence": 0.95, "relation": "introduces", "releasedAt": "2026-09-02", "source": {"authors": ["Matteo Greco", "Anudeex Shetty", "Andrea Tagarelli", "Jey Han Lau"], "categories": ["cs.CL", "cs.AI", "cs.CY", "cs.DL", "cs.IR"], "id": "2609.02379", "title": "MultiGhostBench: A Multilingual Benchmark for Long-Form LLM-Generated Text Attribution under Distribution Shifts", "type": "arxiv", "url": "https://arxiv.org/abs/2609.02379"}, "sourceUpdatedAt": "2026-09-02T09:52:45Z", "topics": ["cs.CL"], "venueAttempts": [], "whyItMatters": "Provides multilingual, long-form data for detecting AI-generated text, a critical need as models improve and produce longer outputs."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
