{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2605.25758", "kind": "paper", "url": "https://arxiv.org/abs/2605.25758"}, {"id": "gh:waterwang-001/streamprofilebench", "kind": "repo", "url": "https://github.com/WaterWang-001/StreamProfileBench"}], "authors": ["Sizhe Wang", "Feiyu Duan", "Juelin Wang", "Liwen Zhang", "Zhongyu Wei"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2605.25758", "source_url": "https://arxiv.org/abs/2605.25758"}, "categories": ["General AI", "Language & Knowledge", "cs.CL"], "description": {"en": "StreamProfileBench evaluates LLMs on fine-grained streaming user profiling. Models maintain a rolling persona summary from a stream of user posts and predict which tags from a candidate pool the user will engage with next. Includes over 120,000 posts from 7,000+ users across five Chinese platforms with metrics for recall, novelty, stability, and error rates."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2605.25758", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2605.25758", "title": "StreamProfileBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["Shanghai University of Finance and Economics", "Fudan University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2605.25758", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2605.25758", "source_url": "https://openalex.org/W7162403734"}, "key": "claire-radar:2605.25758", "modality": null, "name": "StreamProfileBench", "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_streamprofilebench_ccd7b568", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "8494d4805b0e1433d38fc7207f27028cebcda8209a956532741346d06de1f037", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2605.25758"}, "publisher": null, "released": "2026-05-25", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2605.25758", "source_url": "https://arxiv.org/abs/2605.25758"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2605-25758", "source": "claire_radar", "source_benchmark_id": "2605.25758", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/WaterWang-001/StreamProfileBench", "githubScope": "benchmark_repo", "githubStars": 2, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2605.25758", "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"}, "capabilities": [], "confidence": "High", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "bb535c7a9f14f0b08f2e3098bc7a1a486aa3380373a7bad8c12425f61406dbd5", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "StreamProfileBench evaluates LLMs on fine-grained streaming user profiling. Models maintain a rolling persona summary from a stream of user posts and predict which tags from a candidate pool the user will engage with next. Includes over 120,000 posts from 7,000+ users across five Chinese platforms with metrics for recall, novelty, stability, and error rates.", "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", "public artifact URL"], "snippet": "To bridge this gap, we introduce StreamProfileBench, a large-scale benchmark for fine-grained streaming user profiling."}, "familyId": "bmf_5cbc6af978f5", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "existing-code-link", "linkedUrl": "https://github.com/WaterWang-001/StreamProfileBench", "url": "https://github.com/WaterWang-001/StreamProfileBench"}, {"basis": "existing-metric-association", "linkedUrl": "https://github.com/WaterWang-001/StreamProfileBench", "url": "https://huggingface.co/papers/2605.25758"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/WaterWang-001/StreamProfileBench", "url": "https://arxiv.org/abs/2605.25758"}, {"basis": "linked-from-report", "linkedUrl": "https://github.com/WaterWang-001/StreamProfileBench.", "url": "https://arxiv.org/abs/2605.25758"}], "scope": "benchmark_repo", "url": "https://github.com/WaterWang-001/StreamProfileBench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_streamprofilebench_ccd7b568", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/WaterWang-001/StreamProfileBench", "data": null, "hfPaper": "https://huggingface.co/papers/2605.25758", "paper": "https://arxiv.org/abs/2605.25758", "pdf": "https://arxiv.org/pdf/2605.25758", "project": null, "report": "https://arxiv.org/abs/2605.25758"}, "metrics": [], "motivation": "Large Language Models (LLMs) have reshaped user profiling, yet current evaluations mainly focus on static data snapshots.", "name": "StreamProfileBench", "oneLine": "StreamProfileBench evaluates LLMs on fine-grained streaming user profiling. Models maintain a rolling persona summary from a stream of user posts and predict which tags from a candidate pool the user will engage with next. Includes over 120,000 posts from 7,000+ users across five Chinese platforms with metrics for recall, novelty, stability, and error rates.", "paperTitle": "StreamProfileBench: A Benchmark for Fine-Grained User Profile Inference in Real-World Streaming Scenarios", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2605.25758", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "StreamProfileBench Team", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://github.com/WaterWang-001/StreamProfileBench"}], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-05-25", "source": {"authors": ["Sizhe Wang", "Feiyu Duan", "Juelin Wang", "Liwen Zhang", "Zhongyu Wei"], "categories": ["cs.CL"], "id": "2605.25758", "title": "StreamProfileBench: A Benchmark for Fine-Grained User Profile Inference in Real-World Streaming Scenarios", "type": "arxiv", "url": "https://arxiv.org/abs/2605.25758"}, "sourceUpdatedAt": "2026-07-26T10:06:15Z", "topics": ["cs.CL"], "venueAttempts": [], "whyItMatters": "Existing user profiling benchmarks use static data, failing to capture real-world streaming UGC and rapidly evolving interests. StreamProfileBench provides a dynamic evaluation that measures plasticity-stability balance, revealing conservative bias in LLMs and enabling practical improvements for personalized systems."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
