{"record": {"aliases": ["Beyond Global Scalars: Synergizing Token-Level Statistics and Deep Semantics for Adversarial AIGC Text Detection"], "artifacts": [{"id": "arxiv:2608.28009", "kind": "paper", "url": "https://arxiv.org/abs/2608.28009"}, {"id": "gh:tencentbac/neurostat", "kind": "repo", "url": "https://github.com/TencentBAC/NeuroStat"}], "authors": ["Peiming Li", "Yifan Wang", "Zhiyuan Hu", "Shiyu Li", "Zheng Wei", "Yang Tang"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.28009", "source_url": "https://arxiv.org/abs/2608.28009"}, "categories": ["General AI", "Vision & 3D", "cs.CL"], "description": {"en": "Evaluates adversarial robustness of machine-generated text detection across 16,000 samples covering token-level to document-level attacks."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.28009", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.28009", "title": "MOSAIC"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.28009", "modality": null, "name": "MOSAIC", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_beyond-global-scalars-synergizing-token-le_0d7c6f3e", "origin_source": "arxiv", "recognition_confidence": "0.7", "record_sha256": "97cfb811b81bb80cd971b055c1a65c35d93eba12eaae66288be73276a40bbe60", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-06T08:46:12.648160Z", "source_url": "https://arxiv.org/abs/2608.28009"}, "publisher": null, "released": "2026-08-28", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.28009", "source_url": "https://arxiv.org/abs/2608.28009"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-28009", "source": "claire_radar", "source_benchmark_id": "2608.28009", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["Beyond Global Scalars: Synergizing Token-Level Statistics and Deep Semantics for Adversarial AIGC Text Detection"], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-09-06", "attemptedAt": "2026-09-06T01:30:14.942019Z", "githubRepo": "https://github.com/TencentBAC/NeuroStat", "githubScope": "benchmark_repo", "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.28009", "observedAt": "2026-09-05T06:23:21.096343Z", "signalStatus": {"githubStars": {"checkedAt": "2026-09-06T01:30:14.942019Z", "state": "unavailable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-06T01:30:14.942019Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The adversarial AIGC detection benchmark targets a high-interest security and LLM evaluation topic with a large sample scale from a major lab, suggesting moderate to high early attention.", "score": 70}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-06T08:46:12.648160Z", "inputHash": "7c5c1ab32bc3d7dd5bd3902855964a9d5278a6661a9bfc364e088d9a4fefd0d5", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "we introduce MOSAIC, a comprehensive adversarial benchmark comprising 16000 samples across a full-granularity attack spectrum", "canonicalNameSource": "abstract", "decisionReason": "Explicitly introduced as a benchmark with a defined attack spectrum and intended for comparing methods; code and data are available.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-06T08:46:12.648160Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates adversarial robustness of machine-generated text detection across 16,000 samples covering token-level to document-level attacks.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "To expose these flaws, we introduce MOSAIC, a comprehensive adversarial benchmark comprising 16000 samples across a full-granularity attack spectrum."}, "familyId": "bmf_8649e8bb3d3b", "firstSeenAt": "2026-08-31", "heat": null, "id": "bm_beyond-global-scalars-synergizing-token-le_0d7c6f3e", "indexedAt": "2026-08-31T01:03:30.163531Z", "industrySectors": [], "links": {"code": "https://github.com/TencentBAC/NeuroStat", "data": null, "paper": "https://arxiv.org/abs/2608.28009", "pdf": "https://arxiv.org/pdf/2608.28009", "project": null, "report": "https://arxiv.org/abs/2608.28009"}, "metrics": [], "motivation": "The rapid evolution of large language models necessitates robust machine-generated text detection.", "name": "MOSAIC", "oneLine": "Evaluates adversarial robustness of machine-generated text detection across 16,000 samples covering token-level to document-level attacks.", "paperTitle": "Beyond Global Scalars: Synergizing Token-Level Statistics and Deep Semantics for Adversarial AIGC Text Detection", "primaryDomain": "General AI", "publication": {"evidence": "Accepted by EMNLP 2026 Findings", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2608.28009", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "EMNLP 2026 Findings", "verifiedAt": "2026-08-31T01:03:30.163531Z"}, "publications": [], "publishers": [{"name": "Tencent BAC", "organizationType": "company-research-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/TencentBAC/NeuroStat"}], "ranking": {"30d": {"confidence": "Low", "coverage": 0.0, "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": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}, "90d": {"confidence": "Low", "coverage": 0.0, "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": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}}, "readiness": "Runnable", "recognitionConfidence": 0.7, "relation": "introduces", "releasedAt": "2026-08-28", "source": {"authors": ["Peiming Li", "Yifan Wang", "Zhiyuan Hu", "Shiyu Li", "Zheng Wei", "Yang Tang"], "categories": ["cs.CL"], "id": "2608.28009", "title": "Beyond Global Scalars: Synergizing Token-Level Statistics and Deep Semantics for Adversarial AIGC Text Detection", "type": "arxiv", "url": "https://arxiv.org/abs/2608.28009"}, "sourceUpdatedAt": "2026-08-28T07:23:58Z", "topics": ["cs.CL"], "venueAttempts": [{"decisionRaw": "Accepted by EMNLP 2026 Findings", "evidence": [{"level": "author-claim", "observedAt": "2026-08-31T01:03:30.163531Z", "rawValue": "Accepted by EMNLP 2026 Findings", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2608.28009"}], "reviewStatus": "accepted", "venueName": "EMNLP 2026 Findings"}], "whyItMatters": "Addresses the lack of adversarial benchmarks for AIGC text detection by exposing vulnerabilities in both statistical and semantic methods, enabling more robust detector development."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
