{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.24145", "kind": "paper", "url": "https://arxiv.org/abs/2608.24145"}, {"id": "gh:skyorca/trustdabench", "kind": "repo", "url": "https://github.com/Skyorca/TrustDABench"}], "authors": ["Boshen Shi", "Yize Liu", "Chen Zhao", "Ce Chi", "Zhendong Wang", "Xing Wang", "Junlan Feng"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.24145", "source_url": "https://arxiv.org/abs/2608.24145"}, "categories": ["General AI", "Robustness", "Safety & Trustworthiness", "cs.CL"], "description": {"en": "Evaluates LLM reliability and robustness for structured data analysis using 2,340 human-verified perturbed instances derived from evidence-path perturbations, scoring refusal behavior and resistance to table representation changes."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.24145", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.24145", "title": "TrustDABench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["China Mobile (China)", "Communication University of China"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2608.24145", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2608.24145", "source_url": "https://openalex.org/W7204245142"}, "key": "claire-radar:2608.24145", "modality": null, "name": "TrustDABench", "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": "true", "origin_record_id": "bm_trustdabench_96f4ae98", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "e4014a456632feb964798c808e0e445ffd51b6b681c4dc45517a256d9c482d4a", "relation": "introduces", "review_state": "rule-auto-admitted", "reviewed_at": "2026-08-31T04:05:24.949457Z", "source_url": "https://arxiv.org/abs/2608.24145"}, "publisher": null, "released": "2026-08-25", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.24145", "source_url": "https://arxiv.org/abs/2608.24145"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-24145", "source": "claire_radar", "source_benchmark_id": "2608.24145", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Safety & Trustworthiness", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/Skyorca/TrustDABench", "githubScope": "benchmark_repo", "githubStars": 0, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.24145", "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": {"checkedAt": "2026-10-04T15:31:48.030916Z", "state": "unavailable"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The structured data analysis reliability theme is broadly relevant, but the lack of confirmed public artifacts may limit early traction.", "score": 42}, "capabilities": ["Robustness"], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-26T06:08:47.245811Z", "inputHash": "6da3ea7b87151b64cc1079f5392ba72799d2e9769264f765b6df36b9bca77888", "model": "deepseek-v4-pro"}, "curation": {"decisionReason": "Automatically admitted from an explicit named benchmark release with evaluation-protocol evidence at or above the configured publication-confidence threshold.", "policy": "source-evidence-v2", "reviewedAt": "2026-08-31T04:05:24.949457Z", "state": "rule-auto-admitted"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates LLM reliability and robustness for structured data analysis using 2,340 human-verified perturbed instances derived from evidence-path perturbations, scoring refusal behavior and resistance to table representation changes.", "displayEligible": true, "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", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce TrustDABench, a benchmark that operationalizes these questions as reliability and robustness."}, "familyId": "bmf_07f6231ee3bd", "firstSeenAt": "2026-08-28", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/Skyorca/TrustDABench", "url": "https://github.com/Skyorca/TrustDABench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/Skyorca/TrustDABench", "url": "https://huggingface.co/papers/2608.24145"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/Skyorca/TrustDABench", "url": "https://arxiv.org/abs/2608.24145"}], "scope": "benchmark_repo", "url": "https://github.com/Skyorca/TrustDABench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_trustdabench_96f4ae98", "indexedAt": "2026-08-28T13:06:54.083176Z", "industrySectors": [], "links": {"code": "https://github.com/Skyorca/TrustDABench", "data": null, "hfPaper": "https://huggingface.co/papers/2608.24145", "paper": "https://arxiv.org/abs/2608.24145", "pdf": "https://arxiv.org/pdf/2608.24145", "project": null, "report": "https://arxiv.org/abs/2608.24145"}, "metrics": [], "motivation": "LLMs are increasingly used to analyze spreadsheets, CSV files, and other structured data, but producing a correct-looking answer is not the same as producing a trustworthy analysis.", "name": "TrustDABench", "oneLine": "Evaluates LLM reliability and robustness for structured data analysis using 2,340 human-verified perturbed instances derived from evidence-path perturbations, scoring refusal behavior and resistance to table representation changes.", "paperTitle": "TrustDABench: Benchmarking Reliability and Robustness of LLMs for Structured Data Analysis", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.24145", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {"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": 0.0, "value": 0}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 489, "score": 23}, "method": "attention-ranking-v13", "rank": 489, "score": 23}}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-08-25", "source": {"authors": ["Boshen Shi", "Yize Liu", "Chen Zhao", "Ce Chi", "Zhendong Wang", "Xing Wang", "Junlan Feng"], "categories": ["cs.CL", "cs.SE"], "id": "2608.24145", "title": "TrustDABench: Benchmarking Reliability and Robustness of LLMs for Structured Data Analysis", "type": "arxiv", "url": "https://arxiv.org/abs/2608.24145"}, "sourceUpdatedAt": "2026-08-25T07:09:01Z", "topics": ["cs.CL"], "venueAttempts": [], "whyItMatters": "Targets a practical gap in trustworthy data analysis where models may produce unsupported or inconsistent results across table forms."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
