{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RG6DV36NBNPWMYDKWPJAUP4W24","short_pith_number":"pith:RG6DV36N","schema_version":"1.0","canonical_sha256":"89bc3aefcd0b5f66606ab3d20a3f96d713c6fd8b5581f21a32b15a8c057aa300","source":{"kind":"arxiv","id":"2505.16135","version":1},"attestation_state":"computed","paper":{"title":"Sudoku-Bench: Evaluating creative reasoning with Sudoku variants","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Edoardo Cetin, Jeffrey Seely, Llion Jones, Tianyu Zhao, Yuki Imajuku","submitted_at":"2025-05-22T02:24:35Z","abstract_excerpt":"Existing reasoning benchmarks for large language models (LLMs) frequently fail to capture authentic creativity, often rewarding memorization of previously observed patterns. We address this shortcoming with Sudoku-Bench, a curated benchmark of challenging and unconventional Sudoku variants specifically selected to evaluate creative, multi-step logical reasoning. Sudoku variants form an unusually effective domain for reasoning research: each puzzle introduces unique or subtly interacting constraints, making memorization infeasible and requiring solvers to identify novel logical breakthroughs (`"},"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":"2505.16135","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-22T02:24:35Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"54c7bc405e367d2006b826934d8a1a95006f60d302d886eef9a96f467af17524","abstract_canon_sha256":"8ae513bfc7574d4cd59f058bfa264ca8d5773258137eddbf9eb977d75e6b67b6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:07:30.467655Z","signature_b64":"hb53MeckinGwEF2T6dYT4gVzdpciRhcUlfR5g+LTt84qD0axSCdJ3H+nC/9FbWxr61jPrXF34By7VCwje5aWDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"89bc3aefcd0b5f66606ab3d20a3f96d713c6fd8b5581f21a32b15a8c057aa300","last_reissued_at":"2026-07-05T11:07:30.467121Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:07:30.467121Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Sudoku-Bench: Evaluating creative reasoning with Sudoku variants","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Edoardo Cetin, Jeffrey Seely, Llion Jones, Tianyu Zhao, Yuki Imajuku","submitted_at":"2025-05-22T02:24:35Z","abstract_excerpt":"Existing reasoning benchmarks for large language models (LLMs) frequently fail to capture authentic creativity, often rewarding memorization of previously observed patterns. We address this shortcoming with Sudoku-Bench, a curated benchmark of challenging and unconventional Sudoku variants specifically selected to evaluate creative, multi-step logical reasoning. Sudoku variants form an unusually effective domain for reasoning research: each puzzle introduces unique or subtly interacting constraints, making memorization infeasible and requiring solvers to identify novel logical breakthroughs (`"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.16135","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/2505.16135/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":"2505.16135","created_at":"2026-07-05T11:07:30.467192+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.16135v1","created_at":"2026-07-05T11:07:30.467192+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.16135","created_at":"2026-07-05T11:07:30.467192+00:00"},{"alias_kind":"pith_short_12","alias_value":"RG6DV36NBNPW","created_at":"2026-07-05T11:07:30.467192+00:00"},{"alias_kind":"pith_short_16","alias_value":"RG6DV36NBNPWMYDK","created_at":"2026-07-05T11:07:30.467192+00:00"},{"alias_kind":"pith_short_8","alias_value":"RG6DV36N","created_at":"2026-07-05T11:07:30.467192+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08317","citing_title":"Blind-Spots-Bench: Evaluating Blind Spots in Multimodal Models","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2607.02491","citing_title":"G-RRM: Guiding Symbolic Solvers with Recurrent Reasoning Models","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2506.21734","citing_title":"Hierarchical Reasoning Model","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09519","citing_title":"Weighted Rules under the Stable Model Semantics","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25166","citing_title":"Training Transformers as a Universal Computer","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02572","citing_title":"On Training Large Language Models for Long-Horizon Tasks: An Empirical Study of Horizon Length","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24","json":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24.json","graph_json":"https://pith.science/api/pith-number/RG6DV36NBNPWMYDKWPJAUP4W24/graph.json","events_json":"https://pith.science/api/pith-number/RG6DV36NBNPWMYDKWPJAUP4W24/events.json","paper":"https://pith.science/paper/RG6DV36N"},"agent_actions":{"view_html":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24","download_json":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24.json","view_paper":"https://pith.science/paper/RG6DV36N","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.16135&json=true","fetch_graph":"https://pith.science/api/pith-number/RG6DV36NBNPWMYDKWPJAUP4W24/graph.json","fetch_events":"https://pith.science/api/pith-number/RG6DV36NBNPWMYDKWPJAUP4W24/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24/action/storage_attestation","attest_author":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24/action/author_attestation","sign_citation":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24/action/citation_signature","submit_replication":"https://pith.science/pith/RG6DV36NBNPWMYDKWPJAUP4W24/action/replication_record"}},"created_at":"2026-07-05T11:07:30.467192+00:00","updated_at":"2026-07-05T11:07:30.467192+00:00"}