{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2605.28721", "kind": "paper", "url": "https://arxiv.org/abs/2605.28721"}, {"id": "hf:forival/livebrowsecomp", "kind": "dataset", "url": "https://huggingface.co/datasets/Forival/LiveBrowseComp"}], "authors": ["HuiMing Fan", "Xiao Wang", "Zheng Chu", "Qianyu Wang", "Zhuoyao Wang", "Ming Liu", "Bing Qin", "XingYu"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2605.28721", "source_url": "https://arxiv.org/abs/2605.28721"}, "categories": ["General AI", "Language & Knowledge", "cs.AI"], "description": {"en": "A deep-search evaluation containing 335 human-authored questions about recently published, non-globally-salient facts."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2605.28721", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2605.28721", "title": "LiveBrowseComp"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Harbin Institute of Technology", "Xiaohongshu (Little Red Book) (China)"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2605.28721", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2605.28721", "source_url": "https://openalex.org/W7162685862"}, "key": "claire-radar:2605.28721", "modality": null, "name": "LiveBrowseComp", "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": "2026-09-25T04:40:22Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_livebrowsecomp_4dea44d6", "origin_source": "arxiv", "recognition_confidence": "0.8", "record_sha256": "762afa9a6671e58f1ba3263f4bbb555dede6bccaa1f84c09424e486e30d82fa7", "relation": "introduces", "review_model": "codex-cli-default", "review_state": "ai-reviewed", "reviewed_at": "2026-09-24T16:16:12.630337Z", "source_url": "https://arxiv.org/abs/2605.28721"}, "publisher": null, "released": "2026-05-27", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2605.28721", "source_url": "https://arxiv.org/abs/2605.28721"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2605-28721", "source": "claire_radar", "source_benchmark_id": "2605.28721", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "The search-agent focus and public dataset may attract near-term evaluation interest, although future attention cannot be established from the supplied text.", "score": 61}, "capabilities": [], "confidence": "Medium", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-24T16:16:12.630337Z", "inputHash": "17a0c5e0baec7590fbb1c36d84f169f5a5bc14e36502109fe61644a32966e322", "model": "codex-cli-default"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "LiveBrowseComp: Are Search Agents Searching, or Just Verifying What They Already Know?", "canonicalNameSource": "paper_title", "decisionReason": "The source formally names the benchmark, defines its evaluation task, reports comparable accuracy-based results, and provides a public dataset path.", "evaluationEvidence": {"reuse": "LiveBrowseComp is available at https://huggingface.co/datasets/Forival/LiveBrowseComp.", "scoring": "all evaluated agents fall below 2% closed-book accuracy, search-augmented scores drop by 25-40 points relative to BrowseComp", "task": "a deep-search benchmark designed to evaluate agents beyond intrinsic coverage"}, "model": "codex-cli-default", "reviewedAt": "2026-09-24T16:16:12.630337Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "A deep-search evaluation containing 335 human-authored questions about recently published, non-globally-salient facts.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "public artifact URL"], "snippet": "We then introduce LiveBrowseComp, a deep-search benchmark designed to evaluate agents beyond intrinsic coverage."}, "familyId": "bmf_1e3b6f356d48", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_livebrowsecomp_4dea44d6", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": "https://huggingface.co/datasets/Forival/LiveBrowseComp", "paper": "https://arxiv.org/abs/2605.28721", "pdf": "https://arxiv.org/pdf/2605.28721", "project": null, "report": "https://arxiv.org/abs/2605.28721"}, "metrics": [], "motivation": "Are LLM-based search agents genuinely searching, or using the web to verify what they already know?", "name": "LiveBrowseComp", "oneLine": "A deep-search evaluation containing 335 human-authored questions about recently published, non-globally-salient facts.", "paperTitle": "LiveBrowseComp: Are Search Agents Searching, or Just Verifying What They Already Know?", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2605.28721", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "readiness": "Inspectable", "recognitionConfidence": 0.8, "relation": "introduces", "releasedAt": "2026-05-27", "reviewLifecycle": {"state": "pending"}, "source": {"authors": ["HuiMing Fan", "Xiao Wang", "Zheng Chu", "Qianyu Wang", "Zhuoyao Wang", "Ming Liu", "Bing Qin", "XingYu"], "categories": ["cs.AI"], "id": "2605.28721", "title": "LiveBrowseComp: Are Search Agents Searching, or Just Verifying What They Already Know?", "type": "arxiv", "url": "https://arxiv.org/abs/2605.28721"}, "sourceUpdatedAt": "2026-05-27T16:39:57Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "It tests search-agent performance where answers are unlikely to be recoverable from the model's pre-existing knowledge."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
