{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:WA3S3WIY4S75FZJJORLM6UMZVY","short_pith_number":"pith:WA3S3WIY","schema_version":"1.0","canonical_sha256":"b0372dd918e4bfd2e5297456cf5199ae3bf9b0f01248ad71d0d0e7f7c1b8325d","source":{"kind":"arxiv","id":"2302.07867","version":5},"attestation_state":"computed","paper":{"title":"Learning Performance-Improving Code Edits","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.PF"],"primary_cat":"cs.SE","authors_text":"Alexander Shypula, Aman Madaan, Amir Yazdanbakhsh, Graham Neubig, Jacob Gardner, Milad Hashemi, Osbert Bastani, Parthasarathy Ranganathan, Uri Alon, Yimeng Zeng","submitted_at":"2023-02-15T18:59:21Z","abstract_excerpt":"With the decline of Moore's law, optimizing program performance has become a major focus of software research. However, high-level optimizations such as API and algorithm changes remain elusive due to the difficulty of understanding the semantics of code. Simultaneously, pretrained large language models (LLMs) have demonstrated strong capabilities at solving a wide range of programming tasks. To that end, we introduce a framework for adapting LLMs to high-level program optimization. First, we curate a dataset of performance-improving edits made by human programmers of over 77,000 competitive C"},"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":"2302.07867","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2023-02-15T18:59:21Z","cross_cats_sorted":["cs.AI","cs.LG","cs.PF"],"title_canon_sha256":"f7b009f1fa713acad788b61801cfeb1ab17b7fe025c00a1526e430f6affaf371","abstract_canon_sha256":"de0b37906471f24a6ebf7a0c72b87b59ea48c2454e0fb6fde8083b4420ffe498"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:12:25.413180Z","signature_b64":"4vIm2TUwBrDucx6GfUNQgx2wmvnfh+XLrkzxXP1tWMRzmVOt0W7V8zucDibQvLlNc9cgII2MotOjNmAVHrClDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b0372dd918e4bfd2e5297456cf5199ae3bf9b0f01248ad71d0d0e7f7c1b8325d","last_reissued_at":"2026-07-05T08:12:25.412686Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:12:25.412686Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Performance-Improving Code Edits","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","cs.PF"],"primary_cat":"cs.SE","authors_text":"Alexander Shypula, Aman Madaan, Amir Yazdanbakhsh, Graham Neubig, Jacob Gardner, Milad Hashemi, Osbert Bastani, Parthasarathy Ranganathan, Uri Alon, Yimeng Zeng","submitted_at":"2023-02-15T18:59:21Z","abstract_excerpt":"With the decline of Moore's law, optimizing program performance has become a major focus of software research. However, high-level optimizations such as API and algorithm changes remain elusive due to the difficulty of understanding the semantics of code. Simultaneously, pretrained large language models (LLMs) have demonstrated strong capabilities at solving a wide range of programming tasks. To that end, we introduce a framework for adapting LLMs to high-level program optimization. First, we curate a dataset of performance-improving edits made by human programmers of over 77,000 competitive C"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.07867","kind":"arxiv","version":5},"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/2302.07867/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":"2302.07867","created_at":"2026-07-05T08:12:25.412746+00:00"},{"alias_kind":"arxiv_version","alias_value":"2302.07867v5","created_at":"2026-07-05T08:12:25.412746+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.07867","created_at":"2026-07-05T08:12:25.412746+00:00"},{"alias_kind":"pith_short_12","alias_value":"WA3S3WIY4S75","created_at":"2026-07-05T08:12:25.412746+00:00"},{"alias_kind":"pith_short_16","alias_value":"WA3S3WIY4S75FZJJ","created_at":"2026-07-05T08:12:25.412746+00:00"},{"alias_kind":"pith_short_8","alias_value":"WA3S3WIY","created_at":"2026-07-05T08:12:25.412746+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.06821","citing_title":"Chiseling Out Efficiency: Structured Skeleton Supervision for Efficient Code Generation","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31767","citing_title":"JETO-Bench: A Reproducible Benchmark for Execution Time Improvement Patches in Java","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2512.14018","citing_title":"PerfCoder: Large Language Models for Interpretable Code Performance Optimization","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10890","citing_title":"CppPerf: An Automated Pipeline and Dataset for Performance-Improving C++ Commits","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2304.05128","citing_title":"Teaching Large Language Models to Self-Debug","ref_index":109,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23892","citing_title":"Optimas: An Intelligent Analytics-Informed Generative AI Framework for Performance Optimization","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04809","citing_title":"Watts This Smell: A Comprehensive Taxonomy of Software Energy Smells","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2303.17651","citing_title":"Self-Refine: Iterative Refinement with Self-Feedback","ref_index":23,"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":282,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY","json":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY.json","graph_json":"https://pith.science/api/pith-number/WA3S3WIY4S75FZJJORLM6UMZVY/graph.json","events_json":"https://pith.science/api/pith-number/WA3S3WIY4S75FZJJORLM6UMZVY/events.json","paper":"https://pith.science/paper/WA3S3WIY"},"agent_actions":{"view_html":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY","download_json":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY.json","view_paper":"https://pith.science/paper/WA3S3WIY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2302.07867&json=true","fetch_graph":"https://pith.science/api/pith-number/WA3S3WIY4S75FZJJORLM6UMZVY/graph.json","fetch_events":"https://pith.science/api/pith-number/WA3S3WIY4S75FZJJORLM6UMZVY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY/action/storage_attestation","attest_author":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY/action/author_attestation","sign_citation":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY/action/citation_signature","submit_replication":"https://pith.science/pith/WA3S3WIY4S75FZJJORLM6UMZVY/action/replication_record"}},"created_at":"2026-07-05T08:12:25.412746+00:00","updated_at":"2026-07-05T08:12:25.412746+00:00"}