{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.29218", "kind": "paper", "url": "https://arxiv.org/abs/2607.29218"}], "authors": ["Jianxin Gao", "Beini Hu", "Runze Li", "Hongjian Fang", "Wanli Peng"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.29218", "source_url": "https://arxiv.org/abs/2607.29218"}, "categories": ["General AI", "Language & Knowledge", "cs.AI"], "description": {"en": "MirrorCraft is a paired benchmark for evaluating LLM-based agents under hidden rule changes in Minecraft. Each Mirror world copies a Vanilla world with modified server-side rules, while terrain and objectives match. It includes five biomes, six rule suites, three objectives, and uses deterministic milestones and Rule Intervention Effect."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.29218", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.29218", "title": "MirrorCraft"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2607.29218", "modality": null, "name": "MirrorCraft", "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_mirrorcraft_4e4ff3e9", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "5a02cdc16ec93edfff920a4f1583da158b67d447b86283993a4c2af0321fd0bf", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.29218"}, "publisher": null, "released": "2026-07-31", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.29218", "source_url": "https://arxiv.org/abs/2607.29218"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-29218", "source": "claire_radar", "source_benchmark_id": "2607.29218", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "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/2607.29218", "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"}, "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": "047b7389d34807385728a55dbff7e5866c34b714d09e80dc54496f6f82f01002", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "MirrorCraft is a paired benchmark for evaluating LLM-based agents under hidden rule changes in Minecraft. Each Mirror world copies a Vanilla world with modified server-side rules, while terrain and objectives match. It includes five biomes, six rule suites, three objectives, and uses deterministic milestones and Rule Intervention Effect.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "In this paper, we introduce MirrorCraft, a paired benchmark for evaluating agents under hidden rule changes in Minecraft."}, "familyId": "bmf_17f3ccda834b", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 0, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [], "status": "not-found"}, "heat": null, "id": "bm_mirrorcraft_4e4ff3e9", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "hfPaper": "https://huggingface.co/papers/2607.29218", "paper": "https://arxiv.org/abs/2607.29218", "pdf": "https://arxiv.org/pdf/2607.29218", "project": null, "report": "https://arxiv.org/abs/2607.29218"}, "metrics": [], "motivation": "With the prosperity of the large language models (LLMs), it has become an interesting topic: how do LLM-based agents work in Minecraft?", "name": "MirrorCraft", "oneLine": "MirrorCraft is a paired benchmark for evaluating LLM-based agents under hidden rule changes in Minecraft. Each Mirror world copies a Vanilla world with modified server-side rules, while terrain and objectives match. It includes five biomes, six rule suites, three objectives, and uses deterministic milestones and Rule Intervention Effect.", "paperTitle": "MirrorCraft: Paired Evaluation under Hidden Rule Changes in Minecraft", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2607.29218", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [], "publishers": [{"name": "arXiv", "organizationType": "benchmark-organization", "role": "benchmark-publisher", "sourceUrl": "https://arxiv.org/abs/2607.29218"}], "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": "2026-07-31", "source": {"authors": ["Jianxin Gao", "Beini Hu", "Runze Li", "Wanli Peng", "Ruohan Lei", "Jinyuan Zhang", "Linna Deng", "Tianyi Yu", "Zining Wang"], "categories": ["cs.AI"], "id": "2607.29218", "title": "MirrorCraft: Paired Evaluation under Hidden Rule Changes in Minecraft", "type": "arxiv", "url": "https://arxiv.org/abs/2607.29218"}, "sourceUpdatedAt": "2026-07-31T09:43:54Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "Most Minecraft benchmarks use fixed mechanics, not testing adaptability to rule changes. MirrorCraft provides a controlled setting to study agent performance under hidden rule modifications, important for developing agents that can handle dynamic environments."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
