{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.08970", "kind": "paper", "url": "https://arxiv.org/abs/2607.08970"}], "authors": ["Hantao Zhang", "Jinru Sui", "Ed Li", "Dirk Bergemann", "Zhuoran Yang"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.08970", "source_url": "https://arxiv.org/abs/2607.08970"}, "categories": ["General AI", "Geometric reasoning", "Vision & 3D", "cs.CV"], "description": {"en": "Evaluates multi-view integration in vision-language models using diagnostic tasks for 3D scene comprehension, with a fixed-view baseline and proposed ViewNavigator method."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.08970", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.08970", "title": "MultiView-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2607.08970", "modality": null, "name": "MultiView-Bench", "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_multiview-bench_66d1791f", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "9dfb52ee114d0dad5b46524886309261be009ac62faf520637ce8e32e21e907f", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.08970"}, "publisher": null, "released": "2026-07-09", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.08970", "source_url": "https://arxiv.org/abs/2607.08970"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-08970", "source": "claire_radar", "source_benchmark_id": "2607.08970", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 0, "hfPaperUrl": "https://huggingface.co/papers/2607.08970", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "capabilities": ["Geometric reasoning"], "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": "c5f573d94c580572008ee677d38da90719740e72a7bce3aa91d292e048c5ef15", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates multi-view integration in vision-language models using diagnostic tasks for 3D scene comprehension, with a fixed-view baseline and proposed ViewNavigator method.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce MultiView-Bench, a diagnostic benchmark expressly designed to evaluate multi-view integration for holistic 3D scene comprehension."}, "familyId": "bmf_6e0786a96b06", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_multiview-bench_66d1791f", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2607.08970", "paper": "https://arxiv.org/abs/2607.08970", "pdf": "https://arxiv.org/pdf/2607.08970", "project": null, "report": "https://arxiv.org/abs/2607.08970"}, "metrics": [], "motivation": "Recent benchmarks for VLMs largely assess single- or limited-view perception, leaving untested the core cognitive ability to integrate observations across viewpoints into a coherent, world-centric (allocentric) 3D mental model.", "name": "MultiView-Bench", "oneLine": "Evaluates multi-view integration in vision-language models using diagnostic tasks for 3D scene comprehension, with a fixed-view baseline and proposed ViewNavigator method.", "paperTitle": "MultiView-Bench: A Diagnostic Benchmark for World-Centric Multi-View Integration in VLMs", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.08970", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {"90d": {"confidence": "Low", "coverage": 0.15, "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": 0.0, "value": 0}}, "confidence": "Low", "coverage": 0.15, "rank": null, "score": 43}, "method": "attention-ranking-v13", "rank": null, "score": 43}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-07-09", "source": {"authors": ["Hantao Zhang", "Jinru Sui", "Ed Li", "Dirk Bergemann", "Zhuoran Yang"], "categories": ["cs.CV", "cs.AI"], "id": "2607.08970", "title": "MultiView-Bench: A Diagnostic Benchmark for World-Centric Multi-View Integration in VLMs", "type": "arxiv", "url": "https://arxiv.org/abs/2607.08970"}, "sourceUpdatedAt": "2026-08-11T08:09:13Z", "topics": ["cs.CV"], "venueAttempts": [], "whyItMatters": "Addresses the gap in evaluating VLMs' ability to integrate observations across viewpoints into allocentric 3D models, a prerequisite for downstream tasks like assembly."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
