{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QLTVUHNXJQSN3SGJD7AJ64F7IR","short_pith_number":"pith:QLTVUHNX","schema_version":"1.0","canonical_sha256":"82e75a1db74c24ddc8c91fc09f70bf44708f105ea3bbcc694a37b67e9e2e6911","source":{"kind":"arxiv","id":"2507.20395","version":1},"attestation_state":"computed","paper":{"title":"MazeEval: A Benchmark for Testing Sequential Decision-Making in Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Hafsteinn Einarsson","submitted_at":"2025-07-27T19:33:45Z","abstract_excerpt":"As Large Language Models (LLMs) increasingly power autonomous agents in robotics and embodied AI, understanding their spatial reasoning capabilities becomes crucial for ensuring reliable real-world deployment. Despite advances in language understanding, current research lacks evaluation of how LLMs perform spatial navigation without visual cues, a fundamental requirement for agents operating with limited sensory information. This paper addresses this gap by introducing MazeEval, a benchmark designed to isolate and evaluate pure spatial reasoning in LLMs through coordinate-based maze navigation"},"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":"2507.20395","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-27T19:33:45Z","cross_cats_sorted":[],"title_canon_sha256":"cebb27bb8b0b129e09e16765fddf6dec9bd5711400244bcc8f275a7a346df2e5","abstract_canon_sha256":"f1403f0f18eb17db9d7d50d5172f51d01937a430bf3f6fbfaab7c8c0f64bbb71"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:18.532163Z","signature_b64":"D274oUImpAoJ3Zp619at/GfccOOdu4in2ax8U8J/8CJVzbkZH5ZxERDpag88/OJRgd0gCVlSOOj69m99S+ZxAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"82e75a1db74c24ddc8c91fc09f70bf44708f105ea3bbcc694a37b67e9e2e6911","last_reissued_at":"2026-07-05T11:44:18.531708Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:18.531708Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MazeEval: A Benchmark for Testing Sequential Decision-Making in Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Hafsteinn Einarsson","submitted_at":"2025-07-27T19:33:45Z","abstract_excerpt":"As Large Language Models (LLMs) increasingly power autonomous agents in robotics and embodied AI, understanding their spatial reasoning capabilities becomes crucial for ensuring reliable real-world deployment. Despite advances in language understanding, current research lacks evaluation of how LLMs perform spatial navigation without visual cues, a fundamental requirement for agents operating with limited sensory information. This paper addresses this gap by introducing MazeEval, a benchmark designed to isolate and evaluate pure spatial reasoning in LLMs through coordinate-based maze navigation"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.20395","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/2507.20395/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":"2507.20395","created_at":"2026-07-05T11:44:18.531763+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.20395v1","created_at":"2026-07-05T11:44:18.531763+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.20395","created_at":"2026-07-05T11:44:18.531763+00:00"},{"alias_kind":"pith_short_12","alias_value":"QLTVUHNXJQSN","created_at":"2026-07-05T11:44:18.531763+00:00"},{"alias_kind":"pith_short_16","alias_value":"QLTVUHNXJQSN3SGJ","created_at":"2026-07-05T11:44:18.531763+00:00"},{"alias_kind":"pith_short_8","alias_value":"QLTVUHNX","created_at":"2026-07-05T11:44:18.531763+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22219","citing_title":"Lost in Aggregation: A Multi-Scale Diagnostic Benchmark for LLM Spatial Navigation","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28277","citing_title":"Do LLMs Build World Models From Text? A Multilingual Diagnostic of Spatial Reasoning","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09965","citing_title":"Towards Generalist Game Players: An Investigation of Foundation Models in the Game Multiverse","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09965","citing_title":"Towards Generalist Game Players: An Investigation of Foundation Models in the Game Multiverse","ref_index":40,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR","json":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR.json","graph_json":"https://pith.science/api/pith-number/QLTVUHNXJQSN3SGJD7AJ64F7IR/graph.json","events_json":"https://pith.science/api/pith-number/QLTVUHNXJQSN3SGJD7AJ64F7IR/events.json","paper":"https://pith.science/paper/QLTVUHNX"},"agent_actions":{"view_html":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR","download_json":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR.json","view_paper":"https://pith.science/paper/QLTVUHNX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.20395&json=true","fetch_graph":"https://pith.science/api/pith-number/QLTVUHNXJQSN3SGJD7AJ64F7IR/graph.json","fetch_events":"https://pith.science/api/pith-number/QLTVUHNXJQSN3SGJD7AJ64F7IR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR/action/storage_attestation","attest_author":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR/action/author_attestation","sign_citation":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR/action/citation_signature","submit_replication":"https://pith.science/pith/QLTVUHNXJQSN3SGJD7AJ64F7IR/action/replication_record"}},"created_at":"2026-07-05T11:44:18.531763+00:00","updated_at":"2026-07-05T11:44:18.531763+00:00"}