{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2610.09550", "kind": "paper", "url": "https://arxiv.org/abs/2610.09550"}], "authors": ["Huiyao Zhang", "Jin Bai", "Zilong Su", "Rui Guo", "Chaofan Qin", "Jinze Lv", "Wenhui Yu", "Hongfei Wang", "Ye Li"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2610.09550", "source_url": "https://arxiv.org/abs/2610.09550"}, "categories": ["General AI", "Multimodal"], "description": {"en": "Vision-language models increasingly reason through crops, regions, and tool-produced observations."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2610.09550", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2610.09550", "title": "CROSS-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Chinese Academy of Sciences", "Technology and Engineering Center for Space Utilization", "University of Chinese Academy of Sciences", "University of Science and Technology of China"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2610.09550", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2610.09550", "source_url": "https://openalex.org/W7220999379"}, "key": "claire-radar:2610.09550", "modality": null, "name": "CROSS-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": "true", "origin_record_id": "bm_cross-bench_f0bc35a3", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "626332ab2310454225055a5aad6c911b2a21b1496fa24a2b7d444b2cf5ad988e", "relation": "introduces", "review_state": "rule-auto-admitted", "reviewed_at": "2026-10-10T07:28:06.891613Z", "source_url": "https://arxiv.org/abs/2610.09550"}, "publisher": null, "released": "2026-10-07", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2610.09550", "source_url": "https://arxiv.org/abs/2610.09550"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2610-09550", "source": "claire_radar", "source_benchmark_id": "2610.09550", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Multimodal", "capabilities": [], "confidence": "High", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "curation": {"decisionReason": "Automatically admitted from an explicit named benchmark release with evaluation-protocol evidence at or above the configured publication-confidence threshold.", "policy": "source-evidence-v2", "reviewedAt": "2026-10-10T07:28:06.891613Z", "state": "rule-auto-admitted"}, "dataStatus": "primary-source-indexed", "demo": false, "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce CROSS-Bench, a benchmark of 28,000 decision problems, with matched invalidation and rebinding tests on a dedicated evaluation subset."}, "familyId": "bmf_29a6da232923", "firstSeenAt": "2026-10-08", "heat": null, "id": "bm_cross-bench_f0bc35a3", "indexedAt": "2026-10-10T07:27:47.776017Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2610.09550", "pdf": "https://arxiv.org/pdf/2610.09550", "project": null, "report": "https://arxiv.org/abs/2610.09550"}, "metrics": [], "motivation": "Vision-language models increasingly reason through crops, regions, and tool-produced observations.", "name": "CROSS-Bench", "oneLine": "Vision-language models increasingly reason through crops, regions, and tool-produced observations.", "paperTitle": "Visual Evidence Under Cross-Examination: Evaluating and Controlling Decision-Level Evidence Use in Vision-Language Models", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2610.09550", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-10-07", "source": {"authors": ["Huiyao Zhang", "Jin Bai", "Zilong Su", "Rui Guo", "Chaofan Qin", "Jinze Lv", "Wenhui Yu", "Hongfei Wang", "Ye Li"], "categories": ["cs.CV"], "id": "2610.09550", "title": "Visual Evidence Under Cross-Examination: Evaluating and Controlling Decision-Level Evidence Use in Vision-Language Models", "type": "arxiv", "url": "https://arxiv.org/abs/2610.09550"}, "sourceUpdatedAt": "2026-10-07T06:48:44Z", "topics": ["Multimodal"], "venueAttempts": []}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
