{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.06784", "kind": "paper", "url": "https://arxiv.org/abs/2606.06784"}], "authors": ["Zifan Peng", "Yini Huang", "Aiwen Lu", "Qiming Ye", "Peixian Zhang", "Jingyi Zheng", "Yule Liu", "Xuechao Wang", "Xinlei He", "Jiaheng Wei"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.06784", "source_url": "https://arxiv.org/abs/2606.06784"}, "categories": ["General AI", "Language & Knowledge", "cs.CR"], "description": {"en": "SopriBench evaluates user-level privacy leakage from social media posts across text, images, and metadata in 50 synthetic profiles with 1,569 images, scoring via the Privacy Exposure Score (PES) that weights value granularity by contextual sensitivity."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.06784", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.06784", "title": "SopriBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "institutions": ["The Hong Kong University of Science and Technology (Guangzhou)", "Wuhan University"], "institutions_reference": {"basis": "openalex_linked_institutions", "doi": "10.48550/arxiv.2606.06784", "retrieved_at": "2026-11-04", "source_key": "claire-radar:2606.06784", "source_url": "https://openalex.org/W7163923190"}, "key": "claire-radar:2606.06784", "modality": null, "name": "SopriBench", "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_sopribench_12eedd2a", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "f2be3e7dfd9f03b3fbe75bfe4fc04a4fb8c3aab21867e137460e6cdcb12033b5", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.06784"}, "publisher": null, "released": "2026-06-05", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.06784", "source_url": "https://arxiv.org/abs/2606.06784"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-06784", "source": "claire_radar", "source_benchmark_id": "2606.06784", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Machine Generated", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "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/2606.06784", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-09-25", "state": "stale"}}, "status": "stale"}, "capabilities": [], "confidence": "High", "construction": "Original Synthetic", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "b263b18199ab3aad17385941fe9d79d17046bd29c40867517d2eddb4e4064005", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "SopriBench evaluates user-level privacy leakage from social media posts across text, images, and metadata in 50 synthetic profiles with 1,569 images, scoring via the Privacy Exposure Score (PES) that weights value granularity by contextual sensitivity.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact named benchmark artifact released in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "To address these gaps, we propose SopriBench, a synthetic benchmark guided by leakage patterns abstracted from a private reference corpus of Rednote and Instagram accounts, covering 50 user profiles and 1,569 images with attributes, contextual sensitivity, granularity, leakage type, inference difficulty, and supporting evidence."}, "familyId": "bmf_4c1de46427bd", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_sopribench_12eedd2a", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2606.06784", "paper": "https://arxiv.org/abs/2606.06784", "pdf": "https://arxiv.org/pdf/2606.06784", "project": null, "report": "https://arxiv.org/abs/2606.06784"}, "metrics": [], "motivation": "Public social media posts can reveal private information through weak cues scattered across text, images, or metadata.", "name": "SopriBench", "oneLine": "SopriBench evaluates user-level privacy leakage from social media posts across text, images, and metadata in 50 synthetic profiles with 1,569 images, scoring via the Privacy Exposure Score (PES) that weights value granularity by contextual sensitivity.", "paperTitle": "What Your Posts Reveal: A Benchmark and Agentic Framework for User-Level Privacy Leakage on Social Media", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.06784", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-06-05", "source": {"authors": ["Zifan Peng", "Yini Huang", "Aiwen Lu", "Qiming Ye", "Peixian Zhang", "Jingyi Zheng", "Yule Liu", "Xuechao Wang", "Xinlei He", "and Jiaheng Wei"], "categories": ["cs.CR", "cs.AI", "cs.CY"], "id": "2606.06784", "title": "What Your Posts Reveal: A Benchmark and Agentic Framework for User-Level Privacy Leakage on Social Media", "type": "arxiv", "url": "https://arxiv.org/abs/2606.06784"}, "sourceUpdatedAt": "2026-06-05T00:02:47Z", "topics": ["cs.CR"], "venueAttempts": [], "whyItMatters": "This benchmark addresses the gap in evaluating cumulative cross-post privacy leakage, providing a metric that captures exposure severity rather than binary accuracy, aiding in assessing real-world privacy risks from social media data."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
