{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2606.09169", "kind": "paper", "url": "https://arxiv.org/abs/2606.09169"}], "authors": ["Lingyi Meng", "Zecong Tang", "Haoran Li", "Tengju Ru", "Zhejun Cui", "Weitong Lian", "Qi Kang", "Hangshuo Cao", "Yichen Zhu", "Yechi Liu", "Kaixuan Wang", "Yu-Jie Yuan", "Chunwei Wang", "Yu Zhang", "Bo Dai"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.09169", "source_url": "https://arxiv.org/abs/2606.09169"}, "categories": ["General AI", "Multimodal"], "description": {"en": "IMUG-Bench evaluates unified multimodal models on multi-turn interleaved image-text understanding and generation. It includes 3,113 samples and 12,034 interaction turns across Static Spatial, Temporal Causal, and Hybrid classes, with dynamic understanding questions."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.09169", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.09169", "title": "IMUG-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2606.09169", "modality": null, "name": "IMUG-Bench", "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_imug-bench_282ebfce", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "b588c9fc70ac2cd6a33f9da77eaff1fc414aa754091b179abe4eebc1299de056", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2606.09169"}, "publisher": null, "released": "2026-06-08", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.09169", "source_url": "https://arxiv.org/abs/2606.09169"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-09169", "source": "claire_radar", "source_benchmark_id": "2606.09169", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Multimodal", "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/2606.09169", "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": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "ce091803bc0a1f379bbd008d46778b66cba1c3da185d636365d11f760279076c", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "IMUG-Bench evaluates unified multimodal models on multi-turn interleaved image-text understanding and generation. It includes 3,113 samples and 12,034 interaction turns across Static Spatial, Temporal Causal, and Hybrid classes, with dynamic understanding questions.", "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"], "snippet": "To bridge this gap, we propose IMUG-Bench, a comprehensive benchmark for multi-turn interleaved image-text dialogue of UMMs that jointly evaluates their understanding and generation capabilities."}, "familyId": "bmf_92e790a6df66", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_imug-bench_282ebfce", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2606.09169", "paper": "https://arxiv.org/abs/2606.09169", "pdf": "https://arxiv.org/pdf/2606.09169", "project": null, "report": "https://arxiv.org/abs/2606.09169"}, "metrics": [], "motivation": "In recent years, unified multimodal models (UMMs) have emerged to support both understanding and generation within a single framework.", "name": "IMUG-Bench", "oneLine": "IMUG-Bench evaluates unified multimodal models on multi-turn interleaved image-text understanding and generation. It includes 3,113 samples and 12,034 interaction turns across Static Spatial, Temporal Causal, and Hybrid classes, with dynamic understanding questions.", "paperTitle": "IMUG-Bench: Benchmarking Unified Multimodal Models on Interleaved Understanding and Generation", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.09169", "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": "2026-06-08", "source": {"authors": ["Lingyi Meng", "Zecong Tang", "Haoran Li", "Tengju Ru", "Zhejun Cui", "Weitong Lian", "Qi Kang", "Hangshuo Cao", "Yichen Zhu", "Yechi Liu", "Kaixuan Wang", "Yu-Jie Yuan", "Chunwei Wang", "Yu Zhang and Bo Dai"], "categories": ["cs.AI", "cs.CV", "cs.MM"], "id": "2606.09169", "title": "IMUG-Bench: Benchmarking Unified Multimodal Models on Interleaved Understanding and Generation", "type": "arxiv", "url": "https://arxiv.org/abs/2606.09169"}, "sourceUpdatedAt": "2026-06-08T08:08:20Z", "topics": ["Multimodal"], "venueAttempts": [], "whyItMatters": "Existing benchmarks fail to evaluate multi-turn interleaved interactions and expose bias. IMUG-Bench provides a comprehensive evaluation revealing capability boundaries and failure modes, and explores test-time scaling strategies."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
