{"record": {"aliases": ["GUI-Primitives: Diagnosing Spatial Reasoning Failures in Vision-Language GUI Grounding"], "artifacts": [{"id": "arxiv:2608.21832", "kind": "paper", "url": "https://arxiv.org/abs/2608.21832v1"}], "authors": ["Md Abrar Jahin", "Md Rizwan Parvez"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2608.21832", "source_url": "https://arxiv.org/abs/2608.21832"}, "categories": ["General AI", "Multimodal", "Reasoning", "Safety"], "description": {"en": "994-item benchmark of contrastive instruction pairs over seven spatial relations in GUIs, isolating whether vision-language models bind relational language to UI elements correctly."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.21832v1", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.21832v1", "title": "GUI-Primitives"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.21832", "modality": null, "name": "GUI-Primitives", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "Medium", "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_gui-primitives-diagnosing-spatial-reasonin_a22b1551", "origin_source": "arxiv", "recognition_confidence": "0.6", "record_sha256": "4a43851a9d5ff6efc8304a1dd6be5f3bce89264ed1f0300d69daed3a4d479b50", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-08-26T06:15:55.708965Z", "source_url": "https://arxiv.org/abs/2608.21832v1"}, "publisher": null, "released": "2026-08-22", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.21832", "source_url": "https://arxiv.org/abs/2608.21832v1"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-21832", "source": "claire_radar", "source_benchmark_id": "2608.21832", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["GUI-Primitives: Diagnosing Spatial Reasoning Failures in Vision-Language GUI Grounding"], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Multimodal", "attention": {"asOf": "2026-08-26", "attemptedAt": "2026-08-25T17:51:33.814358Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.21832", "observedAt": "2026-08-24T11:31:30.653116Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-08-25T17:51:33.814358Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Relevant to GUI agents and spatial reasoning, but scope may be narrower than general benchmarks.", "score": 45}, "capabilities": ["Reasoning"], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-26T06:20:35.651311Z", "inputHash": "f035db2a2694c4a4cce7c744c092b1ab3d95d40440947a41bd70d5d53e869830", "model": "deepseek-v4-flash"}, "curation": {"canonicalNameEvidence": "GUI-Primitives: Diagnosing Spatial Reasoning Failures in Vision-Language GUI Grounding", "canonicalNameSource": "paper_title", "decisionReason": "The paper introduces a named benchmark with a controlled design, validated annotations, and released code and predictions.", "model": "deepseek-v4-pro", "reviewedAt": "2026-08-26T06:15:55.708965Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "994-item benchmark of contrastive instruction pairs over seven spatial relations in GUIs, isolating whether vision-language models bind relational language to UI elements correctly.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We introduce GUI-Primitives, a 994-item benchmark of contrastive instruction pairs over seven spatial relations in graphical user interfaces (left/right, above/below, containment, alignment, proximity, list ordinal, occlusion)."}, "familyId": "bmf_29e2d419d7ef", "firstSeenAt": "2026-08-25", "heat": null, "id": "bm_gui-primitives-diagnosing-spatial-reasonin_a22b1551", "indexedAt": "2026-08-25T17:37:25.889905Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "http://arxiv.org/abs/2608.21832v1", "pdf": "https://arxiv.org/pdf/2608.21832v1", "project": null, "report": "http://arxiv.org/abs/2608.21832v1"}, "metrics": [], "motivation": "Computer-use agents ground natural-language instructions in screenshots to locate interface elements, yet existing benchmarks do not isolate whether models bind relational language to the correct element.", "name": "GUI-Primitives", "oneLine": "994-item benchmark of contrastive instruction pairs over seven spatial relations in GUIs, isolating whether vision-language models bind relational language to UI elements correctly.", "paperTitle": "GUI-Primitives: Diagnosing Spatial Reasoning Failures in Vision-Language GUI Grounding", "primaryDomain": "General AI", "publication": {"evidence": "Accepted to EMNLP 2026 Main Conference", "evidenceLevel": "author-claim", "evidenceUrl": "http://arxiv.org/abs/2608.21832v1", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "EMNLP 2026 Main Conference", "verifiedAt": "2026-08-25T17:37:25.889905Z"}, "publications": [], "ranking": {"30d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"percentile": null, "value": null}, "hfDatasetDownloads": {"percentile": null, "value": null}, "hfPaperUpvotes": {"percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"percentile": null, "value": null}, "hfDatasetDownloads": {"percentile": null, "value": null}, "hfPaperUpvotes": {"percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v11", "rank": null, "score": null}, "90d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"percentile": null, "value": null}, "hfDatasetDownloads": {"percentile": null, "value": null}, "hfPaperUpvotes": {"percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"percentile": null, "value": null}, "hfDatasetDownloads": {"percentile": null, "value": null}, "hfPaperUpvotes": {"percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v11", "rank": null, "score": null}}, "readiness": "Paper only", "recognitionConfidence": 0.6, "relation": "introduces", "releasedAt": "2026-08-22", "source": {"authors": ["Md Abrar Jahin", "Md Rizwan Parvez"], "categories": ["cs.CL"], "id": "2608.21832", "title": "GUI-Primitives: Diagnosing Spatial Reasoning Failures in Vision-Language GUI Grounding", "type": "arxiv", "url": "http://arxiv.org/abs/2608.21832v1"}, "sourceUpdatedAt": "2026-08-22T08:04:59Z", "topics": ["Multimodal", "Safety", "Reasoning"], "venueAttempts": [{"decisionRaw": "Accepted to EMNLP 2026 Main Conference", "evidence": [{"level": "author-claim", "observedAt": "2026-08-25T17:37:25.889905Z", "rawValue": "Accepted to EMNLP 2026 Main Conference", "sourceType": "arxiv-comments", "sourceUrl": "http://arxiv.org/abs/2608.21832v1"}], "reviewStatus": "accepted", "venueName": "EMNLP 2026 Main Conference"}], "whyItMatters": "Provides fine-grained diagnostics for spatial reasoning in GUI grounding, a key capability gap in computer-use agents, with public code and predictions."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
