{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2608.17600", "kind": "paper", "url": "https://arxiv.org/abs/2608.17600"}], "authors": ["Zhengyan Qian", "Rui Yan", "Alex Jinpeng Wang", "Jinhui Tang"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.17600", "source_url": "https://arxiv.org/abs/2608.17600"}, "categories": ["Embodied manipulation", "Multimodal", "Robotics & Embodied AI", "Safety", "Safety & Trustworthiness", "Visual instruction following"], "description": {"en": "LIBERO-VIFO is a benchmark for visual cue following in vision-language-action models, evaluating capability and safety across eight cue families. However, no public artifacts or official links are provided."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.17600", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.17600", "title": "LIBERO-VIFO"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.17600", "modality": null, "name": "LIBERO-VIFO", "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_libero-vifo_76763b3b", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "7594bc828a1826ef098572c6d577fc86bbe47811d61a1337dd52db911a0ae138", "relation": "introduces", "review_state": "source-reviewed", "reviewed_at": "2026-08-20", "source_url": "https://arxiv.org/abs/2608.17600"}, "publisher": null, "released": "2026-08-18", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.17600", "source_url": "https://arxiv.org/abs/2608.17600"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-17600", "source": "claire_radar", "source_benchmark_id": "2608.17600", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["Robotics & Embodied AI"], "area": "Safety & Trustworthiness", "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": null, "hfPaperUrl": "https://huggingface.co/papers/2608.17600", "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", "state": "unavailable"}}, "status": "unavailable"}, "availability": {"evaluatorStatus": "not_found", "githubStatus": "not_found", "hfDatasetStatus": "not_found", "paperStatus": "available", "submissionStatus": "not_found"}, "capabilities": ["Embodied manipulation", "Visual instruction following", "Safety"], "confidence": "High", "construction": "Unknown", "constructionDetail": "LIBERO-VIFO tests authorized and unauthorized visual influence across robot-manipulation tasks and eight families of visual cues.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "299804e0fc9f5979754ed09eb8b0c5057dff36aab38ecfdc75b78d811ac14934", "model": "deepseek-v4-flash"}, "curation": {"reviewedAt": "2026-08-20", "sources": ["https://arxiv.org/abs/2608.17600", "https://arxiv.org/html/2608.17600"], "state": "source-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "LIBERO-VIFO is a benchmark for visual cue following in vision-language-action models, evaluating capability and safety across eight cue families. However, no public artifacts or official links are provided.", "detail": {"protocol": {"primaryMetric": "Full-Chain Accuracy and authorized/unauthorized visual following rates", "tasks": "1,347 instances from 40 LIBERO tasks and 33 cue variants", "version": "arXiv v1"}, "taskBreakdown": ["Text overlays", "Geometric symbols", "Ghost trajectories", "Pictorial demonstrations", "Hand gestures", "Visual signage", "Visual highlights", "Machine-readable codes"]}, "domainCuration": {"method": "primary-source-review-v1", "state": "reviewed"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "To address these gaps, we introduce LIBERO-VIFO, a benchmark to evaluate both the capability and safety of visual cue following in VLA models."}, "familyId": "bmf_092e28009a80", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_libero-vifo_76763b3b", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": ["Robotics"], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2608.17600", "paper": "https://arxiv.org/abs/2608.17600", "pdf": "https://arxiv.org/pdf/2608.17600", "project": null, "report": "https://arxiv.org/abs/2608.17600"}, "metrics": [], "motivation": "Visual cues are increasingly adopted to guide robot learning, but whether Vision-Language-Action (VLA) models can reliably follow authorized cues while disregarding unauthorized ones remains unclear.", "name": "LIBERO-VIFO", "oneLine": "Evaluates whether vision-language-action models follow authorized visual cues while resisting unauthorized visual influence.", "paperTitle": "LIBERO-VIFO: Benchmarking the Capability and Safety of Visual Cue Following in Vision-Language-Action Models", "primaryDomain": "Robotics & Embodied AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.17600", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {"90d": {"confidence": "Low", "coverage": 0.0, "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": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-08-18", "source": {"authors": ["Zhengyan Qian", "Rui Yan", "Alex Jinpeng Wang", "Jinhui Tang"], "categories": ["cs.RO"], "id": "2608.17600", "title": "LIBERO-VIFO: Benchmarking the Capability and Safety of Visual Cue Following in Vision-Language-Action Models", "type": "arxiv", "url": "https://arxiv.org/abs/2608.17600"}, "sourceUpdatedAt": "2026-08-18T10:07:17Z", "topics": ["Multimodal", "Safety"], "venueAttempts": [], "whyItMatters": "If released, it could inform the development of safer VLA systems, but current lack of access limits its practical value."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
