{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:KMA45QH5JMSNL4LBL6L3DUJVIR","short_pith_number":"pith:KMA45QH5","canonical_record":{"source":{"id":"2502.03568","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T19:30:28Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"12c32a441046d146d30bb7966bcb66f0cfedb52790cfc88eb9ad595f7b48918b","abstract_canon_sha256":"d2068f89276c7b083937d99f1240a245949b1f48b1d90a8cb2032ba130cde1b7"},"schema_version":"1.0"},"canonical_sha256":"5301cec0fd4b24d5f1615f97b1d13544434ad8e0e4972272a52318fc25a54d6f","source":{"kind":"arxiv","id":"2502.03568","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.03568","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"arxiv_version","alias_value":"2502.03568v3","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.03568","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"pith_short_12","alias_value":"KMA45QH5JMSN","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"pith_short_16","alias_value":"KMA45QH5JMSNL4LB","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"pith_short_8","alias_value":"KMA45QH5","created_at":"2026-07-05T11:31:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:KMA45QH5JMSNL4LBL6L3DUJVIR","target":"record","payload":{"canonical_record":{"source":{"id":"2502.03568","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T19:30:28Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"12c32a441046d146d30bb7966bcb66f0cfedb52790cfc88eb9ad595f7b48918b","abstract_canon_sha256":"d2068f89276c7b083937d99f1240a245949b1f48b1d90a8cb2032ba130cde1b7"},"schema_version":"1.0"},"canonical_sha256":"5301cec0fd4b24d5f1615f97b1d13544434ad8e0e4972272a52318fc25a54d6f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:31:43.010960Z","signature_b64":"yoME4R0N0X8WcZfQoRek8cmH2ukuC6URNJoKWbMXWgsPdjqtNNNHBLeFLZqs6G4u6UM9IezkFneBGb5BZajyAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5301cec0fd4b24d5f1615f97b1d13544434ad8e0e4972272a52318fc25a54d6f","last_reissued_at":"2026-07-05T11:31:43.010495Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:31:43.010495Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.03568","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:31:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZuDnFnudoAOrAL6jDb88Bww4mxXPVjHcpRYeqnjQoO7tVZsk5cSpBwtK+8ju0051tZ1diT4FVKm9sqnroZhCDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T03:04:38.453932Z"},"content_sha256":"9fa216ad0f48e7b3170271a42095b1a4a406dec41516ca6fd950539ce3b782e6","schema_version":"1.0","event_id":"sha256:9fa216ad0f48e7b3170271a42095b1a4a406dec41516ca6fd950539ce3b782e6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:KMA45QH5JMSNL4LBL6L3DUJVIR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Code Simulation as a Proxy for High-order Tasks in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Anthony Cohn, Christoph Weinhuber, Emanuele La Malfa, Fangru Lin, Michael Wooldridge, Nigel Shadbolt, Orazio Torre, Samuele Marro, X. Angelo Huang","submitted_at":"2025-02-05T19:30:28Z","abstract_excerpt":"Many reasoning, planning, and problem-solving tasks share an intrinsic algorithmic nature: correctly simulating each step is a sufficient condition to solve them correctly. We collect pairs of naturalistic and synthetic reasoning tasks to assess the capabilities of Large Language Models (LLM). While naturalistic tasks often require careful human handcrafting, we show that synthetic data is, in many cases, a good proxy that is much easier to collect at scale. We leverage common constructs in programming as the counterpart of the building blocks of naturalistic reasoning tasks, such as straight-"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.03568","kind":"arxiv","version":3},"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/2502.03568/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:31:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7h0E5Xcbex/a+fEiq4Aa6bNgbDufKdWZ8nfycmZtyD13dI52hwWuSSrK9C9Xa++IGwRO41vxR1leAhwXjlUmBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T03:04:38.454978Z"},"content_sha256":"c8577ea0a57fe3fd66b5cd89a17238567c84911b6976f55dce86dbb0964f5d47","schema_version":"1.0","event_id":"sha256:c8577ea0a57fe3fd66b5cd89a17238567c84911b6976f55dce86dbb0964f5d47"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/KMA45QH5JMSNL4LBL6L3DUJVIR/bundle.json","state_url":"https://pith.science/pith/KMA45QH5JMSNL4LBL6L3DUJVIR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/KMA45QH5JMSNL4LBL6L3DUJVIR/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-09T03:04:38Z","links":{"resolver":"https://pith.science/pith/KMA45QH5JMSNL4LBL6L3DUJVIR","bundle":"https://pith.science/pith/KMA45QH5JMSNL4LBL6L3DUJVIR/bundle.json","state":"https://pith.science/pith/KMA45QH5JMSNL4LBL6L3DUJVIR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/KMA45QH5JMSNL4LBL6L3DUJVIR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:KMA45QH5JMSNL4LBL6L3DUJVIR","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"d2068f89276c7b083937d99f1240a245949b1f48b1d90a8cb2032ba130cde1b7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T19:30:28Z","title_canon_sha256":"12c32a441046d146d30bb7966bcb66f0cfedb52790cfc88eb9ad595f7b48918b"},"schema_version":"1.0","source":{"id":"2502.03568","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.03568","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"arxiv_version","alias_value":"2502.03568v3","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.03568","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"pith_short_12","alias_value":"KMA45QH5JMSN","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"pith_short_16","alias_value":"KMA45QH5JMSNL4LB","created_at":"2026-07-05T11:31:43Z"},{"alias_kind":"pith_short_8","alias_value":"KMA45QH5","created_at":"2026-07-05T11:31:43Z"}],"graph_snapshots":[{"event_id":"sha256:c8577ea0a57fe3fd66b5cd89a17238567c84911b6976f55dce86dbb0964f5d47","target":"graph","created_at":"2026-07-05T11:31:43Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.03568/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Many reasoning, planning, and problem-solving tasks share an intrinsic algorithmic nature: correctly simulating each step is a sufficient condition to solve them correctly. We collect pairs of naturalistic and synthetic reasoning tasks to assess the capabilities of Large Language Models (LLM). While naturalistic tasks often require careful human handcrafting, we show that synthetic data is, in many cases, a good proxy that is much easier to collect at scale. We leverage common constructs in programming as the counterpart of the building blocks of naturalistic reasoning tasks, such as straight-","authors_text":"Anthony Cohn, Christoph Weinhuber, Emanuele La Malfa, Fangru Lin, Michael Wooldridge, Nigel Shadbolt, Orazio Torre, Samuele Marro, X. Angelo Huang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T19:30:28Z","title":"Code Simulation as a Proxy for High-order Tasks in Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.03568","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:9fa216ad0f48e7b3170271a42095b1a4a406dec41516ca6fd950539ce3b782e6","target":"record","created_at":"2026-07-05T11:31:43Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"d2068f89276c7b083937d99f1240a245949b1f48b1d90a8cb2032ba130cde1b7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-05T19:30:28Z","title_canon_sha256":"12c32a441046d146d30bb7966bcb66f0cfedb52790cfc88eb9ad595f7b48918b"},"schema_version":"1.0","source":{"id":"2502.03568","kind":"arxiv","version":3}},"canonical_sha256":"5301cec0fd4b24d5f1615f97b1d13544434ad8e0e4972272a52318fc25a54d6f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5301cec0fd4b24d5f1615f97b1d13544434ad8e0e4972272a52318fc25a54d6f","first_computed_at":"2026-07-05T11:31:43.010495Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:31:43.010495Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"yoME4R0N0X8WcZfQoRek8cmH2ukuC6URNJoKWbMXWgsPdjqtNNNHBLeFLZqs6G4u6UM9IezkFneBGb5BZajyAw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:31:43.010960Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.03568","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9fa216ad0f48e7b3170271a42095b1a4a406dec41516ca6fd950539ce3b782e6","sha256:c8577ea0a57fe3fd66b5cd89a17238567c84911b6976f55dce86dbb0964f5d47"],"state_sha256":"5f318e3e9fbd6a957571058fc50642f4d7f868dd7d7c11ebb21270f3d51adf5a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"460dvRonwrCoqmMY4nbowOz1AGrN/irhvTYLdr/Wk5YRO6E98UBH2YijTWT7IoxD0+EfhxKwvJbb8cTY5UZ8DQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T03:04:38.465710Z","bundle_sha256":"b6a99a2ee8be800e910ab22b701fa2233e72eeec1320b9155d3524497346820d"}}