{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2605.21919", "kind": "paper", "url": "https://arxiv.org/abs/2605.21919"}], "categories": ["General AI", "Multimodal", "Reasoning"], "description": {"en": "SDGBiasBench is a benchmark suite for evaluating vision-language models on Sustainable Development Goals (SDG) reasoning, covering 500k multiple-choice questions and 50k regression tasks."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2605.21919", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2605.21919", "title": "SDGBiasBench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2605.21919", "modality": null, "name": "SDGBiasBench", "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_sdgbiasbench_aa7777ae", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "c584d960aa894020c3da899d8b37a3a774740aacea7322a7cc4a44021027b7da", "relation": "aggregates", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2605.21919"}, "publisher": null, "released": "2026-05-21", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2605.21919", "source_url": "https://arxiv.org/abs/2605.21919"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2605-21919", "source": "claire_radar", "source_benchmark_id": "2605.21919", "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/2605.21919", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}}, "status": "fresh"}, "capabilities": ["Reasoning"], "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": "6dc257974a66feb822577a03a7b8377bdfba5ddf19c19f1218ec7b61d143cf99", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "SDGBiasBench is a benchmark suite for evaluating vision-language models on Sustainable Development Goals (SDG) reasoning, covering 500k multiple-choice questions and 50k regression tasks.", "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 address this gap, we propose SDGBiasBench, a large-scale benchmark suite for SDG-oriented vision-language reasoning."}, "familyId": "bmf_7819f7afec07", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_sdgbiasbench_aa7777ae", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2605.21919", "paper": "https://arxiv.org/abs/2605.21919", "pdf": "https://arxiv.org/pdf/2605.21919", "project": null, "report": "https://arxiv.org/abs/2605.21919"}, "metrics": [], "motivation": "Assessing progress toward the Sustainable Development Goals (SDGs) requires multi-step reasoning over visual cues, contextual knowledge, and development indicators, where incomplete evidence use and imperfect evidence integration can introduce hidden prediction biases.", "name": "SDGBiasBench", "oneLine": "SDGBiasBench is a benchmark suite for evaluating vision-language models on Sustainable Development Goals (SDG) reasoning, covering 500k multiple-choice questions and 50k regression tasks.", "paperTitle": "SDGBiasBench: Benchmarking and Mitigating Vision--Language Models' Biases in Sustainable Development Goals", "primaryDomain": "General AI", "publication": {"evidence": "Accepted to EMNLP 2026 Findings", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2605.21919", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "EMNLP 2026 Findings", "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "ranking": {}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "aggregates", "releasedAt": "2026-05-21", "source": {"authors": ["Zihang Lin", "Huaiyuan Qin", "Muli Yang", "Hongyuan Zhu"], "categories": ["cs.CV", "cs.AI"], "id": "2605.21919", "title": "SDGBiasBench: Benchmarking and Mitigating Vision--Language Models' Biases in Sustainable Development Goals", "type": "arxiv", "url": "https://arxiv.org/abs/2605.21919"}, "sourceUpdatedAt": "2026-05-21T02:44:54Z", "topics": ["Multimodal", "Reasoning"], "venueAttempts": [{"decisionRaw": "Accepted to EMNLP 2026 Findings", "evidence": [{"level": "author-claim", "observedAt": "2026-10-14T01:46:58.058199Z", "rawValue": "Accepted to EMNLP 2026 Findings", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2605.21919"}], "reviewStatus": "accepted", "venueName": "EMNLP 2026 Findings"}], "whyItMatters": "Existing SDG evaluation tools lack a combined assessment of qualitative and quantitative reasoning, and this benchmark aims to expose systematic biases in model predictions. The value lies in enabling more reliable AI for sustainable development monitoring."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
