{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FPPS6UOI2OOGUMOMT2IP46ZX3Z","short_pith_number":"pith:FPPS6UOI","schema_version":"1.0","canonical_sha256":"2bdf2f51c8d39c6a31cc9e90fe7b37de4abfa9a2b7b49815946264b955b36023","source":{"kind":"arxiv","id":"2607.29218","version":1},"attestation_state":"computed","paper":{"title":"MirrorCraft: Paired Evaluation under Hidden Rule Changes in Minecraft","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Beini Hu, Jianxin Gao, Jinyuan Zhang, Linna Deng, Runze Li, Ruohan Lei, Tianyi Yu, Wanli Peng, Zining Wang","submitted_at":"2026-07-31T09:43:54Z","abstract_excerpt":"With the prosperity of the large language models (LLMs), it has become an interesting topic: how do LLM-based agents work in Minecraft? Unfortunately, most existing benchmarks evaluate them under fixed game mechanics. High performance in these settings does not show whether an agent can continue making progress when familiar recipes, drops, and other rules change. In this paper, we introduce MirrorCraft, a paired benchmark for evaluating agents under hidden rule changes in Minecraft. Each Mirror world is a copy of its paired Vanilla world, with selected server-side rules modified by the corres"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.29218","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-31T09:43:54Z","cross_cats_sorted":[],"title_canon_sha256":"88a9e01a09c09218d7212e50b36c1f1411cddfd75c392295ed9b0063da635c96","abstract_canon_sha256":"d927967ae0183f966f4aa86edd5154828126d8e4530086a50683bd615306db46"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-03T01:21:54.384423Z","signature_b64":"Z7RyTgqebY3O4iVc/tS5NpXXHF0ZVu+S2CVZyurtGbd9/z9GUgATjbdNjPoJiixjGTZm0gsqHsRsImUzbzjJCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2bdf2f51c8d39c6a31cc9e90fe7b37de4abfa9a2b7b49815946264b955b36023","last_reissued_at":"2026-08-03T01:21:54.382662Z","signature_status":"signed_v1","first_computed_at":"2026-08-03T01:21:54.382662Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MirrorCraft: Paired Evaluation under Hidden Rule Changes in Minecraft","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Beini Hu, Jianxin Gao, Jinyuan Zhang, Linna Deng, Runze Li, Ruohan Lei, Tianyi Yu, Wanli Peng, Zining Wang","submitted_at":"2026-07-31T09:43:54Z","abstract_excerpt":"With the prosperity of the large language models (LLMs), it has become an interesting topic: how do LLM-based agents work in Minecraft? Unfortunately, most existing benchmarks evaluate them under fixed game mechanics. High performance in these settings does not show whether an agent can continue making progress when familiar recipes, drops, and other rules change. In this paper, we introduce MirrorCraft, a paired benchmark for evaluating agents under hidden rule changes in Minecraft. Each Mirror world is a copy of its paired Vanilla world, with selected server-side rules modified by the corres"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.29218","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2607.29218/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.29218","created_at":"2026-08-03T01:21:54.383670+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.29218v1","created_at":"2026-08-03T01:21:54.383670+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.29218","created_at":"2026-08-03T01:21:54.383670+00:00"},{"alias_kind":"pith_short_12","alias_value":"FPPS6UOI2OOG","created_at":"2026-08-03T01:21:54.383670+00:00"},{"alias_kind":"pith_short_16","alias_value":"FPPS6UOI2OOGUMOM","created_at":"2026-08-03T01:21:54.383670+00:00"},{"alias_kind":"pith_short_8","alias_value":"FPPS6UOI","created_at":"2026-08-03T01:21:54.383670+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z","json":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z.json","graph_json":"https://pith.science/api/pith-number/FPPS6UOI2OOGUMOMT2IP46ZX3Z/graph.json","events_json":"https://pith.science/api/pith-number/FPPS6UOI2OOGUMOMT2IP46ZX3Z/events.json","paper":"https://pith.science/paper/FPPS6UOI"},"agent_actions":{"view_html":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z","download_json":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z.json","view_paper":"https://pith.science/paper/FPPS6UOI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.29218&json=true","fetch_graph":"https://pith.science/api/pith-number/FPPS6UOI2OOGUMOMT2IP46ZX3Z/graph.json","fetch_events":"https://pith.science/api/pith-number/FPPS6UOI2OOGUMOMT2IP46ZX3Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z/action/storage_attestation","attest_author":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z/action/author_attestation","sign_citation":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z/action/citation_signature","submit_replication":"https://pith.science/pith/FPPS6UOI2OOGUMOMT2IP46ZX3Z/action/replication_record"}},"created_at":"2026-08-03T01:21:54.383670+00:00","updated_at":"2026-08-03T01:21:54.383670+00:00"}