{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.03611", "kind": "paper", "url": "https://arxiv.org/abs/2609.03611"}], "authors": ["Zaruhi Navasardyan", "Tatul Danielyan", "Hrant Davtyan"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.03611", "source_url": "https://arxiv.org/abs/2609.03611"}, "categories": ["General AI", "Robot manipulation", "Robotics", "Robotics & Embodied AI"], "description": {"en": "A robot failure detection benchmark with 2,197 manipulation attempts from real and simulated sources."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.03611", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.03611", "title": "FailBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2609.03611", "modality": null, "name": "FailBench", "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_failbench_15b91781", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "8321b6fa34c326e1a7f6c352555eed8ba88b1fb23c2704086596669220e0b564", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-05T00:38:19.144303Z", "source_url": "https://arxiv.org/abs/2609.03611"}, "publisher": null, "released": "2026-09-03", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.03611", "source_url": "https://arxiv.org/abs/2609.03611"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-03611", "source": "claire_radar", "source_benchmark_id": "2609.03611", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI", "Robotics & Embodied AI"], "area": "Robotics & Embodied AI", "attention": {"asOf": "2026-09-04", "attemptedAt": "2026-09-04T06:36:46.783164Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.03611", "observedAt": "2026-09-03T06:34:20.331209Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-04T06:36:46.783164Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Robot failure detection with VLMs is an active applied area, though missing artifact links in provided metadata may dampen immediate attention.", "score": 55}, "capabilities": ["Robot manipulation"], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-05T00:38:19.144303Z", "inputHash": "a21d120780a647a7c11a58f0fd1509e8f3c36222a822f2cdb2402b78479eb2f0", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "We introduce FailBench, a benchmark for robot failure detection comprising 2,197 manipulation attempts across 14 public sources", "canonicalNameSource": "abstract", "decisionReason": "Formal named benchmark with defined task and cross-domain evaluation.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-05T00:38:19.144303Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "A robot failure detection benchmark with 2,197 manipulation attempts from real and simulated 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", "evaluation protocol evidence"], "snippet": "We introduce FailBench, a benchmark for robot failure detection comprising 2,197 manipulation attempts across 14 public sources (12 real-world, 2 simulated)."}, "familyId": "bmf_e6e85b941f55", "firstSeenAt": "2026-09-04", "heat": null, "id": "bm_failbench_15b91781", "indexedAt": "2026-09-04T06:13:54.443310Z", "industrySectors": ["Robotics"], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.03611", "pdf": "https://arxiv.org/pdf/2609.03611", "project": null, "report": "https://arxiv.org/abs/2609.03611"}, "metrics": [], "motivation": "Vision-Language Models (VLMs) are increasingly used to evaluate robot manipulation outcomes, but existing benchmarks offer limited evidence of cross-domain generalization.", "name": "FailBench", "oneLine": "A robot failure detection benchmark with 2,197 manipulation attempts from real and simulated sources.", "paperTitle": "FailBench: How Reliable are VLMs at Judging Robot Task Success?", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.03611", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-04T06:13:54.443310Z"}, "publications": [], "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}, "today": {"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": 1, "historyDate": "2026-09-03", "rank": null, "score": null, "windowDays": 1}, "level": {"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}, "llmAttentionForecast": {"maxContribution": 0.25, "percentile": 0.4, "role": "bonus", "value": 40}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": 10}, "method": "attention-ranking-v13", "rank": null, "score": 10}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-09-03", "source": {"authors": ["Zaruhi Navasardyan", "Tatul Danielyan", "Hrant Davtyan"], "categories": ["cs.RO", "cs.AI"], "id": "2609.03611", "title": "FailBench: How Reliable are VLMs at Judging Robot Task Success?", "type": "arxiv", "url": "https://arxiv.org/abs/2609.03611"}, "sourceUpdatedAt": "2026-09-03T09:58:55Z", "topics": ["Robotics"], "venueAttempts": [], "whyItMatters": "Reveals limited cross-domain generalization of VLM judges and biased success predictions in ambiguous cases."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
