{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:BMZTDA5MVL6MXI6NPHT4I3Q2BY","short_pith_number":"pith:BMZTDA5M","schema_version":"1.0","canonical_sha256":"0b333183acaafccba3cd79e7c46e1a0e340fb077b642f44b0ed4e9c498763c2d","source":{"kind":"arxiv","id":"2402.11892","version":2},"attestation_state":"computed","paper":{"title":"Towards Reliable Evaluation of Neural Program Repair with Natural Robustness Testing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Bach Le, Dat Nguyen, Thanh Le-Cong, Toby Murray","submitted_at":"2024-02-19T07:07:44Z","abstract_excerpt":"In this paper, we propose shifting the focus of robustness evaluation for Neural Program Repair (NPR) techniques toward naturally-occurring data transformations. To accomplish this, we first examine the naturalness of semantic-preserving transformations through a two-stage human study. This study includes (1) interviews with senior software developers to establish concrete criteria for evaluating the naturalness of these transformations, and (2) a survey involving 10 developers to assess the naturalness of 1,178 transformations, i.e., pairs of original and transformed programs, applied to 225 "},"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":"2402.11892","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-02-19T07:07:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5a0aaee614e0a4cf7e7e4a35ebcc3ee8b793a63699f7e1c5d7de7fd578a7cd52","abstract_canon_sha256":"84980fce55565b826dd35728d27cf4525bd0e93385c2260345306458b29c1db1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:34:34.501280Z","signature_b64":"IXg/p+EzQ/4MTh/fXyg2Wb9adGd7zQON2pKu7QYxfW/Ba3UNtQI1iQ6TnEZ68SqxlmotU6QHMbfYFYZ5gspPDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0b333183acaafccba3cd79e7c46e1a0e340fb077b642f44b0ed4e9c498763c2d","last_reissued_at":"2026-07-05T09:34:34.500884Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:34:34.500884Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Reliable Evaluation of Neural Program Repair with Natural Robustness Testing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Bach Le, Dat Nguyen, Thanh Le-Cong, Toby Murray","submitted_at":"2024-02-19T07:07:44Z","abstract_excerpt":"In this paper, we propose shifting the focus of robustness evaluation for Neural Program Repair (NPR) techniques toward naturally-occurring data transformations. To accomplish this, we first examine the naturalness of semantic-preserving transformations through a two-stage human study. This study includes (1) interviews with senior software developers to establish concrete criteria for evaluating the naturalness of these transformations, and (2) a survey involving 10 developers to assess the naturalness of 1,178 transformations, i.e., pairs of original and transformed programs, applied to 225 "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.11892","kind":"arxiv","version":2},"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/2402.11892/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":"2402.11892","created_at":"2026-07-05T09:34:34.500939+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.11892v2","created_at":"2026-07-05T09:34:34.500939+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.11892","created_at":"2026-07-05T09:34:34.500939+00:00"},{"alias_kind":"pith_short_12","alias_value":"BMZTDA5MVL6M","created_at":"2026-07-05T09:34:34.500939+00:00"},{"alias_kind":"pith_short_16","alias_value":"BMZTDA5MVL6MXI6N","created_at":"2026-07-05T09:34:34.500939+00:00"},{"alias_kind":"pith_short_8","alias_value":"BMZTDA5M","created_at":"2026-07-05T09:34:34.500939+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY","json":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY.json","graph_json":"https://pith.science/api/pith-number/BMZTDA5MVL6MXI6NPHT4I3Q2BY/graph.json","events_json":"https://pith.science/api/pith-number/BMZTDA5MVL6MXI6NPHT4I3Q2BY/events.json","paper":"https://pith.science/paper/BMZTDA5M"},"agent_actions":{"view_html":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY","download_json":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY.json","view_paper":"https://pith.science/paper/BMZTDA5M","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.11892&json=true","fetch_graph":"https://pith.science/api/pith-number/BMZTDA5MVL6MXI6NPHT4I3Q2BY/graph.json","fetch_events":"https://pith.science/api/pith-number/BMZTDA5MVL6MXI6NPHT4I3Q2BY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY/action/storage_attestation","attest_author":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY/action/author_attestation","sign_citation":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY/action/citation_signature","submit_replication":"https://pith.science/pith/BMZTDA5MVL6MXI6NPHT4I3Q2BY/action/replication_record"}},"created_at":"2026-07-05T09:34:34.500939+00:00","updated_at":"2026-07-05T09:34:34.500939+00:00"}