{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:XD7NJ3YU6JM4HUK262QCGOUC54","short_pith_number":"pith:XD7NJ3YU","canonical_record":{"source":{"id":"2401.09074","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-01-17T09:23:59Z","cross_cats_sorted":["cs.AI","cs.CL","cs.PL"],"title_canon_sha256":"fc876c4a3587b612b861bd5d4e74066bf86b78334513b0dcf457d2a561bda18f","abstract_canon_sha256":"176fdfcf8b883a0beac98031542999d923a60c8bdb68fb49173a678768e41e15"},"schema_version":"1.0"},"canonical_sha256":"b8fed4ef14f259c3d15af6a0233a82ef3c44079459290e4b4920f74662bc704e","source":{"kind":"arxiv","id":"2401.09074","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.09074","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"arxiv_version","alias_value":"2401.09074v4","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.09074","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"pith_short_12","alias_value":"XD7NJ3YU6JM4","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"pith_short_16","alias_value":"XD7NJ3YU6JM4HUK2","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"pith_short_8","alias_value":"XD7NJ3YU","created_at":"2026-07-05T08:30:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:XD7NJ3YU6JM4HUK262QCGOUC54","target":"record","payload":{"canonical_record":{"source":{"id":"2401.09074","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-01-17T09:23:59Z","cross_cats_sorted":["cs.AI","cs.CL","cs.PL"],"title_canon_sha256":"fc876c4a3587b612b861bd5d4e74066bf86b78334513b0dcf457d2a561bda18f","abstract_canon_sha256":"176fdfcf8b883a0beac98031542999d923a60c8bdb68fb49173a678768e41e15"},"schema_version":"1.0"},"canonical_sha256":"b8fed4ef14f259c3d15af6a0233a82ef3c44079459290e4b4920f74662bc704e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:30:31.149705Z","signature_b64":"2yXxDIE6a00s4+GXwuA65Pyh7JbJlXPi2V2l0Z2+DKAKaF/6oV7dcN76mhOddXQB8e6bWVY7fqF9VAhcq0aQBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b8fed4ef14f259c3d15af6a0233a82ef3c44079459290e4b4920f74662bc704e","last_reissued_at":"2026-07-05T08:30:31.149131Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:30:31.149131Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2401.09074","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:30:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NlyyLRKr0/lsy733zIaFscM/BfiiYI30mJGsMU6dTLtvKPMgyCAOP0j23g9anp8dtUK97TgU4b+joVli243dCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T12:34:16.879460Z"},"content_sha256":"7b494c173701a729d4220355ae1dfa6d19ad99191819121f5cf51278939c2f64","schema_version":"1.0","event_id":"sha256:7b494c173701a729d4220355ae1dfa6d19ad99191819121f5cf51278939c2f64"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:XD7NJ3YU6JM4HUK262QCGOUC54","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Code Simulation Challenges for Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.PL"],"primary_cat":"cs.LG","authors_text":"Anthony Cohn, Christoph Weinhuber, Emanuele La Malfa, Fangru Lin, Michael Wooldridge, Nigel Shadbolt, Orazio Torre, Samuele Marro","submitted_at":"2024-01-17T09:23:59Z","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. This work studies to what extent Large Language Models (LLMs) can simulate coding and algorithmic tasks to provide insights into general capabilities in such algorithmic reasoning tasks. We introduce benchmarks for straight-line programs, code that contains critical paths, and approximate and redundant instructions. We further assess the simulation capabilities of LLMs with sorting algorithms and nested loops and show that "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.09074","kind":"arxiv","version":4},"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/2401.09074/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":"2026-07-05T08:30:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"enVYSVVett4qxiGLkZmw5ruI34ooKwyNje5sIMXLlKqTF1fvyXzxKgoLkRPaGVGBeAkjCUPD+xgykelFPJqWAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T12:34:16.879950Z"},"content_sha256":"a29845a691234987149f655e01dc5770c13b71b9475e471b2569e4497b0237ec","schema_version":"1.0","event_id":"sha256:a29845a691234987149f655e01dc5770c13b71b9475e471b2569e4497b0237ec"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XD7NJ3YU6JM4HUK262QCGOUC54/bundle.json","state_url":"https://pith.science/pith/XD7NJ3YU6JM4HUK262QCGOUC54/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XD7NJ3YU6JM4HUK262QCGOUC54/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":"2026-08-08T12:34:16Z","links":{"resolver":"https://pith.science/pith/XD7NJ3YU6JM4HUK262QCGOUC54","bundle":"https://pith.science/pith/XD7NJ3YU6JM4HUK262QCGOUC54/bundle.json","state":"https://pith.science/pith/XD7NJ3YU6JM4HUK262QCGOUC54/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XD7NJ3YU6JM4HUK262QCGOUC54/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:XD7NJ3YU6JM4HUK262QCGOUC54","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":"176fdfcf8b883a0beac98031542999d923a60c8bdb68fb49173a678768e41e15","cross_cats_sorted":["cs.AI","cs.CL","cs.PL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-01-17T09:23:59Z","title_canon_sha256":"fc876c4a3587b612b861bd5d4e74066bf86b78334513b0dcf457d2a561bda18f"},"schema_version":"1.0","source":{"id":"2401.09074","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.09074","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"arxiv_version","alias_value":"2401.09074v4","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.09074","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"pith_short_12","alias_value":"XD7NJ3YU6JM4","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"pith_short_16","alias_value":"XD7NJ3YU6JM4HUK2","created_at":"2026-07-05T08:30:31Z"},{"alias_kind":"pith_short_8","alias_value":"XD7NJ3YU","created_at":"2026-07-05T08:30:31Z"}],"graph_snapshots":[{"event_id":"sha256:a29845a691234987149f655e01dc5770c13b71b9475e471b2569e4497b0237ec","target":"graph","created_at":"2026-07-05T08:30:31Z","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/2401.09074/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. This work studies to what extent Large Language Models (LLMs) can simulate coding and algorithmic tasks to provide insights into general capabilities in such algorithmic reasoning tasks. We introduce benchmarks for straight-line programs, code that contains critical paths, and approximate and redundant instructions. We further assess the simulation capabilities of LLMs with sorting algorithms and nested loops and show that ","authors_text":"Anthony Cohn, Christoph Weinhuber, Emanuele La Malfa, Fangru Lin, Michael Wooldridge, Nigel Shadbolt, Orazio Torre, Samuele Marro","cross_cats":["cs.AI","cs.CL","cs.PL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-01-17T09:23:59Z","title":"Code Simulation Challenges for Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.09074","kind":"arxiv","version":4},"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":"sha256:7b494c173701a729d4220355ae1dfa6d19ad99191819121f5cf51278939c2f64","target":"record","created_at":"2026-07-05T08:30:31Z","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":"176fdfcf8b883a0beac98031542999d923a60c8bdb68fb49173a678768e41e15","cross_cats_sorted":["cs.AI","cs.CL","cs.PL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-01-17T09:23:59Z","title_canon_sha256":"fc876c4a3587b612b861bd5d4e74066bf86b78334513b0dcf457d2a561bda18f"},"schema_version":"1.0","source":{"id":"2401.09074","kind":"arxiv","version":4}},"canonical_sha256":"b8fed4ef14f259c3d15af6a0233a82ef3c44079459290e4b4920f74662bc704e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b8fed4ef14f259c3d15af6a0233a82ef3c44079459290e4b4920f74662bc704e","first_computed_at":"2026-07-05T08:30:31.149131Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:30:31.149131Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"2yXxDIE6a00s4+GXwuA65Pyh7JbJlXPi2V2l0Z2+DKAKaF/6oV7dcN76mhOddXQB8e6bWVY7fqF9VAhcq0aQBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:30:31.149705Z","signed_message":"canonical_sha256_bytes"},"source_id":"2401.09074","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7b494c173701a729d4220355ae1dfa6d19ad99191819121f5cf51278939c2f64","sha256:a29845a691234987149f655e01dc5770c13b71b9475e471b2569e4497b0237ec"],"state_sha256":"53a06ab14ff4a87f7039d2a80edc1c776c398ecbeb107e49d4cbcddfd9927f9d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Krskxvw/EwqnkJb+k3R9SNw64XTzjRhnK++lAoE5LSMPj13Ny2yqgXdEYB2CraVsSfdYT/YOCLs67WxKjDy5BQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T12:34:16.885118Z","bundle_sha256":"611fab7762027caa0b013665536faabf8c365fb11b7a2611da1f76d9a173d354"}}