{"record": {"aliases": ["MMPCBench: Benchmarking Multimodal Large Language Models on Proactive Critique of Flawed Inputs"], "artifacts": [{"id": "arxiv:2608.29286", "kind": "paper", "url": "https://arxiv.org/abs/2608.29286"}, {"id": "gh:aliens32/mmpcbench", "kind": "repo", "url": "https://github.com/ALIENS32/MMPCBench"}], "categories": ["General AI", "Multimodal"], "description": {"en": "Evaluates multimodal large language models on proactive critique across 3,146 instances with a taxonomy of 4 error types and 12 subtypes, measuring detection, diagnosis, and resolution using a hierarchical protocol."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2608.29286", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2608.29286", "title": "MMPCBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2608.29286", "modality": null, "name": "MMPCBench", "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_mmpcbench-benchmarking-multimodal-large-la_afb36b3a", "origin_source": "arxiv", "recognition_confidence": "0.55", "record_sha256": "86977cf4ff0c6145d17c49182ec95eabfac39b30b049b6ea1553bb7abc48ee64", "relation": "evaluates_only", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-02T05:23:11.715633Z", "source_url": "https://arxiv.org/abs/2608.29286"}, "publisher": null, "released": "2026-08-29", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2608.29286", "source_url": "https://arxiv.org/abs/2608.29286"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2608-29286", "source": "claire_radar", "source_benchmark_id": "2608.29286", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["MMPCBench: Benchmarking Multimodal Large Language Models on Proactive Critique of Flawed Inputs"], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Multimodal", "attention": {"asOf": "2026-09-02", "attemptedAt": "2026-09-02T05:12:51.621571Z", "githubRepo": "https://github.com/ALIENS32/MMPCBench", "githubScope": "benchmark_repo", "githubStars": 0, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2608.29286", "observedAt": "2026-09-02T05:12:51.621571Z", "signalStatus": {"githubStars": {"observedAt": "2026-09-02T05:12:51.621571Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-02T05:12:51.621571Z", "state": "unavailable"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The benchmark fills a growing interest in MLLM input validation, and the public repository with evaluation pipeline and results is likely to attract immediate use.", "score": 71}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-02T05:23:11.715633Z", "inputHash": "239b59cb9f6c9ce8ded03222d278f9c5096272a2749df72373545fe35fecc09b", "model": "deepseek-v4-pro"}, "curation": {"canonicalNameEvidence": "To fill this gap, we propose MMPCBench, a comprehensive framework for evaluating MLLMs' proactive critique competence.", "canonicalNameSource": "abstract", "decisionReason": "Formally named benchmark with a detailed taxonomy, scoring protocol, and public code and dataset links.", "model": "deepseek-v4-pro", "reviewedAt": "2026-09-02T05:23:11.715633Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates multimodal large language models on proactive critique across 3,146 instances with a taxonomy of 4 error types and 12 subtypes, measuring detection, diagnosis, and resolution using a hierarchical protocol.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "evaluation protocol evidence", "public artifact URL", "named entity is a model, method, framework, or dataset used in evaluation"], "snippet": "To fill this gap, we propose MMPCBench, a comprehensive framework for evaluating MLLMs' proactive critique competence."}, "familyId": "bmf_8acc423a0cfe", "firstSeenAt": "2026-09-01", "heat": null, "id": "bm_mmpcbench-benchmarking-multimodal-large-la_afb36b3a", "indexedAt": "2026-09-01T06:28:29.895112Z", "industrySectors": [], "links": {"code": "https://github.com/ALIENS32/MMPCBench", "data": null, "paper": "https://arxiv.org/abs/2608.29286", "pdf": "https://arxiv.org/pdf/2608.29286", "project": null, "report": "https://arxiv.org/abs/2608.29286"}, "metrics": [], "motivation": "As Multimodal Large Language Models (MLLMs) evolve into sophisticated interactive assistants, their reliability depends not only on following instructions but also on validating them.", "name": "MMPCBench", "oneLine": "Evaluates multimodal large language models on proactive critique across 3,146 instances with a taxonomy of 4 error types and 12 subtypes, measuring detection, diagnosis, and resolution using a hierarchical protocol.", "paperTitle": "MMPCBench: Benchmarking Multimodal Large Language Models on Proactive Critique of Flawed Inputs", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2608.29286", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-01T06:28:29.895112Z"}, "publications": [], "publishers": [{"name": "ALIENS32", "organizationType": "community", "role": "benchmark-publisher", "sourceUrl": "https://github.com/ALIENS32/MMPCBench"}], "ranking": {"30d": {"confidence": "Low", "coverage": 0.55, "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": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": 0.0, "value": 0}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 125, "score": 23}, "method": "attention-ranking-v13", "rank": 125, "score": 23}, "90d": {"confidence": "Low", "coverage": 0.55, "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": 0.0, "value": 0}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 318, "score": 23}, "method": "attention-ranking-v13", "rank": 318, "score": 23}}, "readiness": "Runnable", "recognitionConfidence": 0.55, "relation": "evaluates_only", "releasedAt": "2026-08-29", "source": {"authors": ["Jinzhe Li and Gengxu Li and Jinnan Li and Yuan Wu and Yi Chang"], "categories": ["cs.AI"], "id": "2608.29286", "title": "MMPCBench: Benchmarking Multimodal Large Language Models on Proactive Critique of Flawed Inputs", "type": "arxiv", "url": "https://arxiv.org/abs/2608.29286"}, "sourceUpdatedAt": "2026-08-29T14:19:10Z", "topics": ["Multimodal"], "venueAttempts": [], "whyItMatters": "Targets a neglected aspect of MLLM reliability—validating flawed user inputs—and provides reproducible code and a dataset for measuring this capability."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
