{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:CIQGGQI4MU7XBGG4NSMP44GKPD","short_pith_number":"pith:CIQGGQI4","schema_version":"1.0","canonical_sha256":"122063411c653f7098dc6c98fe70ca78c211241649abcf1b4b96cf88cb48e0fe","source":{"kind":"arxiv","id":"2305.05383","version":1},"attestation_state":"computed","paper":{"title":"Code Execution with Pre-trained Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.SE"],"primary_cat":"cs.PL","authors_text":"Alexey Svyatkovskiy, Chenxiao Liu, Daxin Jiang, Nan Duan, Neel Sundaresan, Shengyu Fu, Shuai Lu, Weizhu Chen","submitted_at":"2023-05-08T10:00:05Z","abstract_excerpt":"Code execution is a fundamental aspect of programming language semantics that reflects the exact behavior of the code. However, most pre-trained models for code intelligence ignore the execution trace and only rely on source code and syntactic structures. In this paper, we investigate how well pre-trained models can understand and perform code execution. We develop a mutation-based data augmentation technique to create a large-scale and realistic Python dataset and task for code execution, which challenges existing models such as Codex. We then present CodeExecutor, a Transformer model that le"},"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":"2305.05383","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.PL","submitted_at":"2023-05-08T10:00:05Z","cross_cats_sorted":["cs.AI","cs.CL","cs.SE"],"title_canon_sha256":"b2efaf330e2bb47ec940b9885c3ac7b3bf25070e2913fc9bebc0b6d34b4c4639","abstract_canon_sha256":"6ba27b772ab2ce73b354b2f41973777d743e1bec3d5365e5433fe977d867ded3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:08:34.062780Z","signature_b64":"E46Qa3viMS3LQ3uIPTndNDxfZ2Kigo/Cb/Dn5bVUG4vdhD0VuGKFYeOd1xI37dHQZirE9h/SOBNozQLDc3rOBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"122063411c653f7098dc6c98fe70ca78c211241649abcf1b4b96cf88cb48e0fe","last_reissued_at":"2026-07-05T06:08:34.062214Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:08:34.062214Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Code Execution with Pre-trained Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.SE"],"primary_cat":"cs.PL","authors_text":"Alexey Svyatkovskiy, Chenxiao Liu, Daxin Jiang, Nan Duan, Neel Sundaresan, Shengyu Fu, Shuai Lu, Weizhu Chen","submitted_at":"2023-05-08T10:00:05Z","abstract_excerpt":"Code execution is a fundamental aspect of programming language semantics that reflects the exact behavior of the code. However, most pre-trained models for code intelligence ignore the execution trace and only rely on source code and syntactic structures. In this paper, we investigate how well pre-trained models can understand and perform code execution. We develop a mutation-based data augmentation technique to create a large-scale and realistic Python dataset and task for code execution, which challenges existing models such as Codex. We then present CodeExecutor, a Transformer model that le"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.05383","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/2305.05383/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":"2305.05383","created_at":"2026-07-05T06:08:34.062274+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.05383v1","created_at":"2026-07-05T06:08:34.062274+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.05383","created_at":"2026-07-05T06:08:34.062274+00:00"},{"alias_kind":"pith_short_12","alias_value":"CIQGGQI4MU7X","created_at":"2026-07-05T06:08:34.062274+00:00"},{"alias_kind":"pith_short_16","alias_value":"CIQGGQI4MU7XBGG4","created_at":"2026-07-05T06:08:34.062274+00:00"},{"alias_kind":"pith_short_8","alias_value":"CIQGGQI4","created_at":"2026-07-05T06:08:34.062274+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.14415","citing_title":"SWE-Chain: Benchmarking Coding Agents on Chained Release-Level Package Upgrades","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2401.03065","citing_title":"CRUXEval: A Benchmark for Code Reasoning, Understanding and Execution","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11922","citing_title":"StepCodeReasoner: Aligning Code Reasoning with Stepwise Execution Traces via Reinforcement Learning","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25166","citing_title":"Training Transformers as a Universal Computer","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06184","citing_title":"Teaching LLMs Program Semantics via Symbolic Execution Traces","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2403.07974","citing_title":"LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD","json":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD.json","graph_json":"https://pith.science/api/pith-number/CIQGGQI4MU7XBGG4NSMP44GKPD/graph.json","events_json":"https://pith.science/api/pith-number/CIQGGQI4MU7XBGG4NSMP44GKPD/events.json","paper":"https://pith.science/paper/CIQGGQI4"},"agent_actions":{"view_html":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD","download_json":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD.json","view_paper":"https://pith.science/paper/CIQGGQI4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.05383&json=true","fetch_graph":"https://pith.science/api/pith-number/CIQGGQI4MU7XBGG4NSMP44GKPD/graph.json","fetch_events":"https://pith.science/api/pith-number/CIQGGQI4MU7XBGG4NSMP44GKPD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD/action/storage_attestation","attest_author":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD/action/author_attestation","sign_citation":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD/action/citation_signature","submit_replication":"https://pith.science/pith/CIQGGQI4MU7XBGG4NSMP44GKPD/action/replication_record"}},"created_at":"2026-07-05T06:08:34.062274+00:00","updated_at":"2026-07-05T06:08:34.062274+00:00"}