{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LIYRG6S4TCNLDJ2HOKFAQR4VZI","short_pith_number":"pith:LIYRG6S4","schema_version":"1.0","canonical_sha256":"5a31137a5c989ab1a747728a084795ca224ed3ccd3f1c79288de1cc621b19187","source":{"kind":"arxiv","id":"2605.13280","version":1},"attestation_state":"computed","paper":{"title":"The Readability Spectrum: Patterns, Issues, and Prompt Effects in LLM-Generated Code","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"LLM-generated code matches human-written code in overall readability but exhibits different issue patterns, and prompt engineering has limited impact on improving it.","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Fengyuan Ran, Hengzhi Ye, Minghui Zhou, Weiwei Xu","submitted_at":"2026-05-13T09:58:21Z","abstract_excerpt":"As Large Language Models (LLMs) are transforming software development, the functional quality of generated code has become a central focus, leaving readability, one of critical non-functional attributes, understudied. Given that LLM-generated code still needs human review before adoption, it is important to understand its readability especially compared with human-written code and the role of prompt design in shaping it. We therefore set out to conduct a systematic investigation into the code readability of LLM-generated code. To systematically quantify code readability, We establish a compreh"},"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":true,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.13280","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-05-13T09:58:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"35e8836822ec9828f822e8ed1bf4f045f7011c74490ad06a5fc9e6cbbcb9cda1","abstract_canon_sha256":"c4fce59a1e7b1a69fc8a7b5029f7f235ad67858dad0ba61a5938394363dd870d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:49.203446Z","signature_b64":"EwEKSDegmAPOuv9paJJUMFj603M46f8aSiVnYILyJOtGfGX1gvI/ZFqjHAhG7P7sfHkojxBg+WoN4+R+xk/LCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5a31137a5c989ab1a747728a084795ca224ed3ccd3f1c79288de1cc621b19187","last_reissued_at":"2026-05-18T02:44:49.202967Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:49.202967Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Readability Spectrum: Patterns, Issues, and Prompt Effects in LLM-Generated Code","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"LLM-generated code matches human-written code in overall readability but exhibits different issue patterns, and prompt engineering has limited impact on improving it.","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Fengyuan Ran, Hengzhi Ye, Minghui Zhou, Weiwei Xu","submitted_at":"2026-05-13T09:58:21Z","abstract_excerpt":"As Large Language Models (LLMs) are transforming software development, the functional quality of generated code has become a central focus, leaving readability, one of critical non-functional attributes, understudied. Given that LLM-generated code still needs human review before adoption, it is important to understand its readability especially compared with human-written code and the role of prompt design in shaping it. We therefore set out to conduct a systematic investigation into the code readability of LLM-generated code. To systematically quantify code readability, We establish a compreh"},"claims":{"count":3,"items":[{"kind":"strongest_claim","text":"We find that current LLMs produce code with overall readability comparable to human-written code, but displaying distinct readability issue patterns.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The synthesized readability model combining textual, structural, program, and visual features accurately measures readability in a manner that aligns with human perception and practical maintainability.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"LLM-generated code matches human-written code in overall readability but exhibits different issue patterns, and prompt engineering has limited impact on improving it.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"}],"snapshot_sha256":"c77b2d415086a1edccc749066a79fc25e6396b275bef2f660e9837192ae262a3"},"source":{"id":"2605.13280","kind":"arxiv","version":1},"verdict":{"id":"27605db9-363d-4d26-87c2-06d7e72681b3","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T18:15:20.006987Z","strongest_claim":"We find that current LLMs produce code with overall readability comparable to human-written code, but displaying distinct readability issue patterns.","one_line_summary":"LLM-generated code matches human-written code in overall readability but exhibits different issue patterns, and prompt engineering has limited impact on improving it.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The synthesized readability model combining textual, structural, program, and visual features accurately measures readability in a manner that aligns with human perception and practical maintainability.","pith_extraction_headline":""},"references":{"count":114,"sample":[{"doi":"","year":null,"title":"[n. d.]. LeetCode. https://leetcode.com/problemset/","work_id":"892fcaf5-6c49-4d9c-899b-e80520950d9b","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"GPT-4 Technical Report","work_id":"b928e041-6991-4c08-8c81-0359e4097c7b","ref_index":2,"cited_arxiv_id":"2303.08774","is_internal_anchor":true},{"doi":"","year":2019,"title":"Duaa Alawad, Manisha Panta, Minhaz Zibran, and Md Rakibul Islam. 2019. An empirical study of the relationships between code readability and software complexity.arXiv preprint arXiv:1909.01760(2019)","work_id":"9dcecb86-5fb0-4cf4-9266-8762afc99ffb","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2010,"title":"André Altmann, Laura Toloşi, Oliver Sander, and Thomas Lengauer. 2010. Per- mutation importance: a corrected feature importance measure.Bioinformatics 26, 10 (2010), 1340–1347","work_id":"18bb6ee4-8546-4ef6-8b74-56a8804f43b7","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Anthropic. [n. d.]. Claude. https://www.anthropic.com/claude/","work_id":"e822cf1f-e8f1-4fe4-a71b-b9739795723b","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":114,"snapshot_sha256":"ddd80698704fdcf9d598fa91117379afd710955e1be1c547f7c52cd177f58b38","internal_anchors":10},"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":"2605.13280","created_at":"2026-05-18T02:44:49.203047+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.13280v1","created_at":"2026-05-18T02:44:49.203047+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13280","created_at":"2026-05-18T02:44:49.203047+00:00"},{"alias_kind":"pith_short_12","alias_value":"LIYRG6S4TCNL","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"LIYRG6S4TCNLDJ2H","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"LIYRG6S4","created_at":"2026-05-18T12:33:37.589309+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/LIYRG6S4TCNLDJ2HOKFAQR4VZI","json":"https://pith.science/pith/LIYRG6S4TCNLDJ2HOKFAQR4VZI.json","graph_json":"https://pith.science/api/pith-number/LIYRG6S4TCNLDJ2HOKFAQR4VZI/graph.json","events_json":"https://pith.science/api/pith-number/LIYRG6S4TCNLDJ2HOKFAQR4VZI/events.json","paper":"https://pith.science/paper/LIYRG6S4"},"agent_actions":{"view_html":"https://pith.science/pith/LIYRG6S4TCNLDJ2HOKFAQR4VZI","download_json":"https://pith.science/pith/LIYRG6S4TCNLDJ2HOKFAQR4VZI.json","view_paper":"https://pith.science/paper/LIYRG6S4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.13280&json=true","fetch_graph":"https://pith.science/api/pith-number/LIYRG6S4TCNLDJ2HOKFAQR4VZI/graph.json","fetch_events":"https://pith.science/api/pith-number/LIYRG6S4TCNLDJ2HOKFAQR4VZI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LIYRG6S4TCNLDJ2HOKFAQR4VZI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LIYRG6S4TCNLDJ2HOKFAQR4VZI/action/storage_attestation","attest_author":"https://pith.science/pith/LIYRG6S4TCNLDJ2HOKFAQR4VZI/action/author_attestation","sign_citation":"https://pith.science/pith/LIYRG6S4TCNLDJ2HOKFAQR4VZI/action/citation_signature","submit_replication":"https://pith.science/pith/LIYRG6S4TCNLDJ2HOKFAQR4VZI/action/replication_record"}},"created_at":"2026-05-18T02:44:49.203047+00:00","updated_at":"2026-05-18T02:44:49.203047+00:00"}