{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.11515", "kind": "paper", "url": "https://arxiv.org/abs/2609.11515"}, {"id": "gh:vsingh45/churnbench", "kind": "repo", "url": "https://github.com/vsingh45/churnbench"}], "categories": ["Code & Software", "General AI", "cs.SE"], "description": {"en": "Measures staleness errors in agentic AI answers under temporal drift across heterogeneous enterprise data sources."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.11515", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.11515", "title": "ChurnBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2609.11515", "modality": null, "name": "ChurnBench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "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_churnbench_8f71ce1d", "origin_source": "arxiv", "recognition_confidence": "0.95", "record_sha256": "78b39d4ea73c96bb2e4938e5b753cdfb8914df92f0d2b2a5b1c5cf20adfe2042", "relation": "extends", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-12T00:56:20.009268Z", "source_url": "https://arxiv.org/abs/2609.11515"}, "publisher": null, "released": "2026-09-10", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.11515", "source_url": "https://arxiv.org/abs/2609.11515"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-11515", "source": "claire_radar", "source_benchmark_id": "2609.11515", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Code & Software", "attention": {"asOf": "2026-09-11", "attemptedAt": "2026-09-11T06:43:09.447014Z", "githubRepo": "https://github.com/vsingh45/churnbench", "githubScope": "benchmark_repo", "githubStars": 0, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.11515", "observedAt": "2026-09-11T06:43:09.447014Z", "signalStatus": {"githubStars": {"observedAt": "2026-09-11T06:43:09.447014Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-11T06:43:09.447014Z", "state": "unavailable"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Agentic AI benchmarking is active, and the strong reproducibility setup for drift and staleness may attract deployment-focused researchers.", "score": 40}, "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-12T00:56:20.009268Z", "inputHash": "1e016050d11daabb8880a22a245ad883aae76ff1be7f875fe1ef22c23e3c0356", "model": "deepseek-v4-pro"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "We present ChurnBench, an open-source benchmark that generates a four-source enterprise data fabric as a timeline rather than a snapshot.", "canonicalNameSource": "abstract", "decisionReason": "The repository offers a repeatable environment, generated tasks, and a clear scoring method based on ground-truth ledgers.", "evaluationEvidence": {"reuse": "Docker-compose infrastructure, dataset generation, snapshots, task creation, and scoring scripts are publicly available.", "reuse_note": "Paper draft and code in repository; committed result data for verification.", "scoring": "Accuracy against a ground-truth ledger and freshness error measure correct answer at true time vs. effective retrieval time.", "task": "Enterprise data fabric across SQL, NoSQL, REST API, and documents, with queries about freshness across simulated timelines."}, "model": "deepseek-v4-pro", "reviewedAt": "2026-09-12T00:56:20.009268Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Measures staleness errors in agentic AI answers under temporal drift across heterogeneous enterprise data sources.", "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", "public artifact URL"], "snippet": "We present ChurnBench, an open-source benchmark that generates a four-source enterprise data fabric as a timeline rather than a snapshot."}, "familyId": "bmf_63a36a05ab9b", "firstSeenAt": "2026-09-11", "heat": null, "id": "bm_churnbench_8f71ce1d", "indexedAt": "2026-09-11T06:20:12.012226Z", "industrySectors": [], "links": {"code": "https://github.com/vsingh45/churnbench", "data": null, "paper": "https://arxiv.org/abs/2609.11515", "pdf": "https://arxiv.org/pdf/2609.11515", "project": null, "report": "https://arxiv.org/abs/2609.11515"}, "metrics": [], "motivation": "In production, agentic systems answer questions over data that lives in several places and keeps changing: licenses are reassigned, users offboarded, prices changed, contracts renewed.", "name": "ChurnBench", "oneLine": "Measures staleness errors in agentic AI answers under temporal drift across heterogeneous enterprise data sources.", "paperTitle": "ChurnBench: A Drift-Aware Benchmark Demonstrating That Refresh Scheduling, Not Cache Age, Governs Staleness in Agentic AI", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.11515", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-11T06:20:12.012226Z"}, "publications": [], "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": 0.0, "value": 0}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 145, "score": 23}, "method": "attention-ranking-v13", "rank": 145, "score": 23}, "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": 337, "score": 23}, "method": "attention-ranking-v13", "rank": 337, "score": 23}, "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": "2026-09-10", "rank": null, "score": null, "windowDays": 1}, "level": {"components": {"githubStars": {"normalization": "percentile", "normalized": 0.4, "percentile": 0.4, "value": 0}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "llmAttentionForecast": {"maxContribution": 0.25, "percentile": 0.75, "role": "bonus", "value": 75}}, "confidence": "Low", "coverage": 0.25, "rank": 3, "score": 54}, "method": "attention-ranking-v13", "rank": 3, "score": 54}}, "readiness": "Runnable", "recognitionConfidence": 0.95, "relation": "extends", "releasedAt": "2026-09-10", "source": {"authors": ["Vivek Kumar Singh", "Preeti Priyam"], "categories": ["cs.SE"], "id": "2609.11515", "title": "ChurnBench: A Drift-Aware Benchmark Demonstrating That Refresh Scheduling, Not Cache Age, Governs Staleness in Agentic AI", "type": "arxiv", "url": "https://arxiv.org/abs/2609.11515"}, "sourceUpdatedAt": "2026-09-10T13:19:29Z", "topics": ["cs.SE"], "venueAttempts": [], "whyItMatters": "Quantifies freshness error in RAG and agent systems, distinguishing scheduling from cache-age effects on correctness."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
