{"record": {"aliases": ["Can We Trust Video Hallucination Detectors? VidHalLoc for Evaluating the Evaluators"], "artifacts": [{"id": "arxiv:2609.09895", "kind": "paper", "url": "https://arxiv.org/abs/2609.09895"}, {"id": "hf:wesfggfd/vidhalloc", "kind": "dataset", "url": "https://huggingface.co/datasets/wesfggfd/VidHalLoc"}], "authors": ["Xinyu Chen", "Adnan Mahmood", "Mark Dras"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.09895", "source_url": "https://arxiv.org/abs/2609.09895"}, "categories": ["Software & AI Compute", "Vision & 3D", "cs.CV"], "description": {"en": "A benchmark with 2,000 adversarial video hallucination samples for evaluating hallucination detectors."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.09895", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.09895", "title": "VidHalLoc"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2609.09895", "modality": null, "name": "VidHalLoc", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "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_can-we-trust-video-hallucination-detectors_9edaaba4", "origin_source": "arxiv", "recognition_confidence": "0.7", "record_sha256": "24eda7a8fe53e25282d094e23399ff8fc9d854286d71ebf57292744f3179810f", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-10T06:21:31.238194Z", "source_url": "https://arxiv.org/abs/2609.09895"}, "publisher": null, "released": "2026-09-09", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.09895", "source_url": "https://arxiv.org/abs/2609.09895"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-09895", "source": "claire_radar", "source_benchmark_id": "2609.09895", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["Can We Trust Video Hallucination Detectors? VidHalLoc for Evaluating the Evaluators"], "annotation": "Unknown", "applicationDomains": ["Software & AI Compute"], "area": "Vision & 3D", "attention": {"asOf": "2026-09-10", "attemptedAt": "2026-09-10T01:41:07.072503Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": "wesfggfd/VidHalLoc", "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.09895", "observedAt": "2026-09-09T12:31:07.558041Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"checkedAt": "2026-09-10T01:41:07.072503Z", "state": "unavailable"}, "hfDatasetLikes": {"checkedAt": "2026-09-10T01:41:07.072503Z", "state": "unavailable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-10T01:41:07.072503Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "The niche focus on evaluating hallucination detectors in video models may attract limited but targeted interest.", "score": 30}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-10T06:21:31.238194Z", "inputHash": "147fc7210a65b49aa0dd896489513201af3b858c65582ab76c8d1d1a9d2363e6", "model": "deepseek-v4-pro"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "VidHalLoc for Evaluating the Evaluators", "canonicalNameSource": "paper_title", "decisionReason": "The dataset is publicly available on Hugging Face, allowing reuse for benchmarking hallucination detection methods.", "evaluationEvidence": {"reuse": "Dataset Repository: https://huggingface.co/datasets/wesfggfd/VidHalLoc", "scoring": "Overall accuracy of detectors on detecting hallucinations.", "task": "Evaluates hallucination detection methods under a unified diagnostic evaluation protocol using 2,000 adversarial hallucination samples across Video Question Answering and Video Captioning tasks."}, "model": "deepseek-v4-pro", "reviewedAt": "2026-09-10T06:21:31.238194Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "A benchmark with 2,000 adversarial video hallucination samples for evaluating hallucination detectors.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce VidHalLoc, a benchmark that evaluates hallucination detection methods under a unified diagnostic evaluation protocol using 2,000 adversarial hallucination samples across Video Question Answering and Video Captioning tasks, spanning Ontology and Dynamic hallucination categories."}, "familyId": "bmf_64d357f74d16", "firstSeenAt": "2026-09-10", "heat": null, "id": "bm_can-we-trust-video-hallucination-detectors_9edaaba4", "indexedAt": "2026-09-10T00:47:26.889471Z", "industrySectors": ["Software & Cloud"], "links": {"code": null, "data": "https://huggingface.co/datasets/wesfggfd/VidHalLoc", "paper": "https://arxiv.org/abs/2609.09895", "pdf": "https://arxiv.org/pdf/2609.09895", "project": null, "report": "https://arxiv.org/abs/2609.09895"}, "metrics": [], "motivation": "Video-language models and video agents can produce hallucinations that conflict with spatiotemporal evidence.", "name": "VidHalLoc", "oneLine": "A benchmark with 2,000 adversarial video hallucination samples for evaluating hallucination detectors.", "paperTitle": "Can We Trust Video Hallucination Detectors? VidHalLoc for Evaluating the Evaluators", "primaryDomain": "Software & AI Compute", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2609.09895", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-10T00:47:26.889471Z"}, "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": "2026-09-09", "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": 0.55, "role": "bonus", "value": 55}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": 14}, "method": "attention-ranking-v13", "rank": null, "score": 14}}, "readiness": "Inspectable", "recognitionConfidence": 0.7, "relation": "introduces", "releasedAt": "2026-09-09", "source": {"authors": ["Xinyu Chen", "Adnan Mahmood", "Mark Dras"], "categories": ["cs.CV"], "id": "2609.09895", "title": "Can We Trust Video Hallucination Detectors? VidHalLoc for Evaluating the Evaluators", "type": "arxiv", "url": "https://arxiv.org/abs/2609.09895"}, "sourceUpdatedAt": "2026-09-09T08:52:23Z", "topics": ["cs.CV"], "venueAttempts": [], "whyItMatters": "Provides a standardized way to compare video hallucination detectors, revealing their limited reliability."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
