{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YLXYNEMZUAFOBYXR75FCI3XLKJ","short_pith_number":"pith:YLXYNEMZ","schema_version":"1.0","canonical_sha256":"c2ef869199a00ae0e2f1ff4a246eeb526f451c1ecb08c66225062df19ce94f75","source":{"kind":"arxiv","id":"2507.07313","version":1},"attestation_state":"computed","paper":{"title":"Frontier LLMs Still Struggle with Simple Reasoning Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Alan Malek, Andr\\'as Gy\\\"orgy, Chi Jin, Csaba Szepesv\\'ari, Jiawei Ge, Nevena Lazic","submitted_at":"2025-07-09T22:22:49Z","abstract_excerpt":"While state-of-the-art large language models (LLMs) demonstrate advanced reasoning capabilities-achieving remarkable performance on challenging competitive math and coding benchmarks-they also frequently fail on tasks that are easy for humans. This work studies the performance of frontier LLMs on a broad set of such \"easy\" reasoning problems. By extending previous work in the literature, we create a suite of procedurally generated simple reasoning tasks, including counting, first-order logic, proof trees, and travel planning, with changeable parameters (such as document length. or the number o"},"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":"2507.07313","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-09T22:22:49Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"9ab6d82c54367a6f3f9080bcaed81477d3ff07dd06e710ff0ff514e798992138","abstract_canon_sha256":"d4caac9086db6ad8d66e0b03e54246fe861ed9f41b630a41b729f8a86cfadd56"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:34:55.683576Z","signature_b64":"AW+YWOLBokM9DA50QZ+P9ezF9gFDMBAdTc8yWMaHM5n30UQ87ReqAaBIrSpyQDK+I/dPv6Bs3cRhug1yYZhfCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c2ef869199a00ae0e2f1ff4a246eeb526f451c1ecb08c66225062df19ce94f75","last_reissued_at":"2026-07-05T11:34:55.683099Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:34:55.683099Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Frontier LLMs Still Struggle with Simple Reasoning Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Alan Malek, Andr\\'as Gy\\\"orgy, Chi Jin, Csaba Szepesv\\'ari, Jiawei Ge, Nevena Lazic","submitted_at":"2025-07-09T22:22:49Z","abstract_excerpt":"While state-of-the-art large language models (LLMs) demonstrate advanced reasoning capabilities-achieving remarkable performance on challenging competitive math and coding benchmarks-they also frequently fail on tasks that are easy for humans. This work studies the performance of frontier LLMs on a broad set of such \"easy\" reasoning problems. By extending previous work in the literature, we create a suite of procedurally generated simple reasoning tasks, including counting, first-order logic, proof trees, and travel planning, with changeable parameters (such as document length. or the number o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.07313","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/2507.07313/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":"2507.07313","created_at":"2026-07-05T11:34:55.683156+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.07313v1","created_at":"2026-07-05T11:34:55.683156+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.07313","created_at":"2026-07-05T11:34:55.683156+00:00"},{"alias_kind":"pith_short_12","alias_value":"YLXYNEMZUAFO","created_at":"2026-07-05T11:34:55.683156+00:00"},{"alias_kind":"pith_short_16","alias_value":"YLXYNEMZUAFOBYXR","created_at":"2026-07-05T11:34:55.683156+00:00"},{"alias_kind":"pith_short_8","alias_value":"YLXYNEMZ","created_at":"2026-07-05T11:34:55.683156+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.11893","citing_title":"Beyond representational alignment with brain-guided language models for robust reasoning","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22964","citing_title":"Certification from Examples is Hard for Circuits and Transformers under Minimal Overparametrization","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11388","citing_title":"Deep Reasoning in General Purpose Agents via Structured Meta-Cognition","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06799","citing_title":"Beyond Accuracy: Diagnosing Algebraic Reasoning Failures in LLMs Across Nine Complexity Dimensions","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21632","citing_title":"To See the Unseen: on the Generalization Ability of Transformers in Symbolic Reasoning","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ","json":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ.json","graph_json":"https://pith.science/api/pith-number/YLXYNEMZUAFOBYXR75FCI3XLKJ/graph.json","events_json":"https://pith.science/api/pith-number/YLXYNEMZUAFOBYXR75FCI3XLKJ/events.json","paper":"https://pith.science/paper/YLXYNEMZ"},"agent_actions":{"view_html":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ","download_json":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ.json","view_paper":"https://pith.science/paper/YLXYNEMZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.07313&json=true","fetch_graph":"https://pith.science/api/pith-number/YLXYNEMZUAFOBYXR75FCI3XLKJ/graph.json","fetch_events":"https://pith.science/api/pith-number/YLXYNEMZUAFOBYXR75FCI3XLKJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ/action/storage_attestation","attest_author":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ/action/author_attestation","sign_citation":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ/action/citation_signature","submit_replication":"https://pith.science/pith/YLXYNEMZUAFOBYXR75FCI3XLKJ/action/replication_record"}},"created_at":"2026-07-05T11:34:55.683156+00:00","updated_at":"2026-07-05T11:34:55.683156+00:00"}