{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.04587", "kind": "paper", "url": "https://arxiv.org/abs/2608.04587"}, {"id": "gh:alibaba-velldepth/metavideoagent", "kind": "repo", "url": "https://github.com/Alibaba-VELLDEPTH/MetaVideoAgent"}], "categories": ["Agents", "General AI", "Vision & 3D"], "description": {"en": "Evaluates long-form video understanding agents across eight video distributions with separate evolution and held-out splits."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.04587", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.04587", "title": "VA-EvoBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.04587", "modality": null, "name": "VA-EvoBench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"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_va-evobench_24a7c4ca", "origin_source": "arxiv", "recognition_confidence": "0.55", "record_sha256": "d3c76c4198b6857127c151b4c8701843dfabf2def1f298b2870eb535ebd7ecb4", "relation": "evaluates_only", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-07T09:32:11.620138Z", "source_url": "https://arxiv.org/abs/2608.04587"}, "publisher": null, "released": "2026-08-05", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.04587", "source_url": "https://arxiv.org/abs/2608.04587"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-04587", "source": "claire_radar", "source_benchmark_id": "2608.04587", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Mixed", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/Alibaba-VELLDEPTH/MetaVideoAgent", "githubScope": "benchmark_repo", "githubStars": 59, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.04587", "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": "Addresses long-form video understanding with automated agent evolution and includes a released benchmark, likely attracting moderate attention.", "score": 55}, "capabilities": [], "confidence": "Medium", "construction": "Aggregate Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-07T09:32:11.620138Z", "inputHash": "1b9a3c93d9c10fd06bbafa9ba69b742077e65fbc161ef800cb44064ffd0b1b76", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "We further introduce VA-EvoBench, covering eight video distributions with separate evolution and held-out splits.", "canonicalNameSource": "abstract", "decisionReason": "The benchmark is formally named, has defined splits, an evaluator, and is released with code and annotations in the repository.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-07T09:32:11.620138Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates long-form video understanding agents across eight video distributions with separate evolution and held-out splits.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "evaluation protocol evidence", "public artifact URL", "named entity is a model, method, framework, or dataset used in evaluation"], "snippet": "We further introduce VA-EvoBench, covering eight video distributions with separate evolution and held-out splits."}, "familyId": "bmf_e06625246c6a", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_va-evobench_24a7c4ca", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/Alibaba-VELLDEPTH/MetaVideoAgent", "data": null, "paper": "https://arxiv.org/abs/2608.04587", "pdf": "https://arxiv.org/pdf/2608.04587", "project": null, "report": "https://arxiv.org/abs/2608.04587"}, "metrics": [], "motivation": "Long-form video understanding requires locating sparse, question-relevant evidence in long, multimodal videos.", "name": "VA-EvoBench", "oneLine": "Evaluates long-form video understanding agents across eight video distributions with separate evolution and held-out splits.", "paperTitle": "MetaVideoAgent: Automated Video-Agent Evolution for Long-Form Video Understanding", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.04587", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "publishers": [{"name": "Alibaba VELLDEPTH", "organizationType": "company-research-lab", "role": "benchmark-publisher", "sourceUrl": "https://github.com/Alibaba-VELLDEPTH/MetaVideoAgent"}], "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": 0.5465213986229674, "value": 59}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 45, "score": 53}, "method": "attention-ranking-v13", "rank": 45, "score": 53}}, "readiness": "Runnable", "recognitionConfidence": 0.55, "relation": "evaluates_only", "releasedAt": "2026-08-05", "source": {"authors": ["Benlei Cui", "Ruize Wang", "Junjie Li", "Jinhao Chen", "Longtao Huang", "Yinghao Chen", "Yuwen Zhai", "Jingqun Tang", "Ruijian Jia", "Weiwei Wu", "Pengfei Sun", "Haiwen Hong"], "categories": ["cs.CV"], "id": "2608.04587", "title": "MetaVideoAgent: Automated Video-Agent Evolution for Long-Form Video Understanding", "type": "arxiv", "url": "https://arxiv.org/abs/2608.04587"}, "sourceUpdatedAt": "2026-08-05T08:50:23Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "Provides a distribution-diverse evaluation suite that reveals fixed video-agent mismatches and supports automated agent evolution research."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
