{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.15899", "kind": "paper", "url": "https://arxiv.org/abs/2607.15899"}, {"id": "gh:vishal-sys-code/continuity-bench", "kind": "repo", "url": "https://github.com/Vishal-sys-code/continuity-bench"}], "authors": ["Vishal Pandey", "Gopal Singh"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.15899", "source_url": "https://arxiv.org/abs/2607.15899"}, "categories": ["General AI", "Language & Knowledge", "cs.LG"], "description": {"en": "ContinuityBench evaluates stateful failover in multi-provider LLM routing. It measures Continuity Preservation Rate (CPR) and Continuity Latency Overhead (CLO) using synthetic conversation graphs under injected provider failures, with an LLM-as-a-judge scoring context preservation."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.15899", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.15899", "title": "ContinuityBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2607.15899", "modality": null, "name": "ContinuityBench", "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_continuitybench_4edb780a", "origin_source": "arxiv", "recognition_confidence": "0.95", "record_sha256": "36286b2cbaaacea352c00392de34b4723b36dd6f4c26de4c439470aa98a55cff", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.15899"}, "publisher": null, "released": "2026-07-17", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.15899", "source_url": "https://arxiv.org/abs/2607.15899"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-15899", "source": "claire_radar", "source_benchmark_id": "2607.15899", "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/Vishal-sys-code/continuity-bench", "githubScope": "benchmark_repo", "githubStars": 0, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 1, "hfPaperUrl": "https://huggingface.co/papers/2607.15899", "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": [], "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": "e3ed0c74b4f0d991c7b651bbde0bf14acbce3836bf9e7b5e6ffd6db48b55bb9d", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "ContinuityBench evaluates stateful failover in multi-provider LLM routing. It measures Continuity Preservation Rate (CPR) and Continuity Latency Overhead (CLO) using synthetic conversation graphs under injected provider failures, with an LLM-as-a-judge scoring context preservation.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "public artifact URL"], "snippet": "Furthermore, we release continuity-bench, https://github.com/Vishal-sys-code/continuity-bench, an open evaluation harness designed to stress-test context preservation under high-concurrency provider failure conditions."}, "familyId": "bmf_74fd5bbd7858", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/Vishal-sys-code/continuity-bench", "url": "https://github.com/Vishal-sys-code/continuity-bench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/Vishal-sys-code/continuity-bench", "url": "https://huggingface.co/papers/2607.15899"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/Vishal-sys-code/continuity-bench", "url": "https://arxiv.org/abs/2607.15899"}], "scope": "benchmark_repo", "url": "https://github.com/Vishal-sys-code/continuity-bench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_continuitybench_4edb780a", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/Vishal-sys-code/continuity-bench", "data": null, "hfPaper": "https://huggingface.co/papers/2607.15899", "paper": "https://arxiv.org/abs/2607.15899", "pdf": "https://arxiv.org/pdf/2607.15899", "project": null, "report": "https://arxiv.org/abs/2607.15899"}, "metrics": [], "motivation": "In production large language model (LLM) deployments, high API availability guarantees do not equate to conversational continuity.", "name": "ContinuityBench", "oneLine": "ContinuityBench evaluates stateful failover in multi-provider LLM routing. It measures Continuity Preservation Rate (CPR) and Continuity Latency Overhead (CLO) using synthetic conversation graphs under injected provider failures, with an LLM-as-a-judge scoring context preservation.", "paperTitle": "ContinuityBench: A Benchmark and Systems Study of Stateful Failover in Multi-Provider LLM Routing", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.15899", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {"90d": {"confidence": "Medium", "coverage": 0.7, "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": 0.12247537038455211, "value": 1}}, "confidence": "Medium", "coverage": 0.7, "rank": 569, "score": 17}, "method": "attention-ranking-v13", "rank": 569, "score": 17}}, "readiness": "Runnable", "recognitionConfidence": 0.95, "relation": "introduces", "releasedAt": "2026-07-17", "source": {"authors": ["Vishal Pandey", "Gopal Singh"], "categories": ["cs.LG"], "id": "2607.15899", "title": "ContinuityBench: A Benchmark and Systems Study of Stateful Failover in Multi-Provider LLM Routing", "type": "arxiv", "url": "https://arxiv.org/abs/2607.15899"}, "sourceUpdatedAt": "2026-07-17T12:12:38Z", "topics": ["cs.LG"], "venueAttempts": [], "whyItMatters": "Production LLM deployments rely on failover mechanisms that often lose conversational context, degrading user experience. This benchmark provides a standardized way to quantify and compare context preservation and latency trade-offs across different routing architectures."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
