{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:R7IFGXYQNHYNXPGY4TOUWELGB7","short_pith_number":"pith:R7IFGXYQ","schema_version":"1.0","canonical_sha256":"8fd0535f1069f0dbbcd8e4dd4b11660ffaf705cac24e4e0edd5dce2ec4e1fa92","source":{"kind":"arxiv","id":"2606.00920","version":1},"attestation_state":"computed","paper":{"title":"Accuracy, Stability, and Repeated-Run Reliability of Large Language Models on Deterministic Programming Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SE"],"primary_cat":"cs.LG","authors_text":"Jiaxi Wen, Lai Yun Choi, Wenbo Ye, Yongxi Zhou","submitted_at":"2026-05-30T23:03:05Z","abstract_excerpt":"Run-level pass rate overstates retry-free coverage by up to 17.8 percentage points -- and the gap is largest precisely for mid-performing systems. We investigate this accuracy--stability relationship in large language model (LLM) evaluation for deterministic text-conditioned generation, using programming tasks as a concrete testbed. Standard code-generation benchmarks emphasize single-run accuracy or eventual success under repeated sampling, but many deployment settings also require stability: consistent outcomes across repeated invocations under the same task description. We present a repeate"},"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":"2606.00920","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-30T23:03:05Z","cross_cats_sorted":["cs.AI","cs.SE"],"title_canon_sha256":"1aca987d93feb8f0626fcd21efcd660d7ebf2d05f33168199c9d21e17b84a030","abstract_canon_sha256":"b247a0ea8460a1432ac444c4cb5f93742535a6bccb9fd5d78b44763e763d88c8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:04:09.500546Z","signature_b64":"7c06LV3RV5IFq6jG0YAidRGZWi4hSceBdWKiXCkNeoAZD/+lKAxYZnj9aFhANrQKhkTzAXBq2EJSOxEyFLllBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8fd0535f1069f0dbbcd8e4dd4b11660ffaf705cac24e4e0edd5dce2ec4e1fa92","last_reissued_at":"2026-06-02T01:04:09.500139Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:04:09.500139Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Accuracy, Stability, and Repeated-Run Reliability of Large Language Models on Deterministic Programming Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SE"],"primary_cat":"cs.LG","authors_text":"Jiaxi Wen, Lai Yun Choi, Wenbo Ye, Yongxi Zhou","submitted_at":"2026-05-30T23:03:05Z","abstract_excerpt":"Run-level pass rate overstates retry-free coverage by up to 17.8 percentage points -- and the gap is largest precisely for mid-performing systems. We investigate this accuracy--stability relationship in large language model (LLM) evaluation for deterministic text-conditioned generation, using programming tasks as a concrete testbed. Standard code-generation benchmarks emphasize single-run accuracy or eventual success under repeated sampling, but many deployment settings also require stability: consistent outcomes across repeated invocations under the same task description. We present a repeate"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.00920","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/2606.00920/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":"2606.00920","created_at":"2026-06-02T01:04:09.500194+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.00920v1","created_at":"2026-06-02T01:04:09.500194+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.00920","created_at":"2026-06-02T01:04:09.500194+00:00"},{"alias_kind":"pith_short_12","alias_value":"R7IFGXYQNHYN","created_at":"2026-06-02T01:04:09.500194+00:00"},{"alias_kind":"pith_short_16","alias_value":"R7IFGXYQNHYNXPGY","created_at":"2026-06-02T01:04:09.500194+00:00"},{"alias_kind":"pith_short_8","alias_value":"R7IFGXYQ","created_at":"2026-06-02T01:04:09.500194+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.02665","citing_title":"Single Canonical Prompts Underestimate LLM Safety's Surface-Form Sensitivity","ref_index":20,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7","json":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7.json","graph_json":"https://pith.science/api/pith-number/R7IFGXYQNHYNXPGY4TOUWELGB7/graph.json","events_json":"https://pith.science/api/pith-number/R7IFGXYQNHYNXPGY4TOUWELGB7/events.json","paper":"https://pith.science/paper/R7IFGXYQ"},"agent_actions":{"view_html":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7","download_json":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7.json","view_paper":"https://pith.science/paper/R7IFGXYQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.00920&json=true","fetch_graph":"https://pith.science/api/pith-number/R7IFGXYQNHYNXPGY4TOUWELGB7/graph.json","fetch_events":"https://pith.science/api/pith-number/R7IFGXYQNHYNXPGY4TOUWELGB7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7/action/storage_attestation","attest_author":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7/action/author_attestation","sign_citation":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7/action/citation_signature","submit_replication":"https://pith.science/pith/R7IFGXYQNHYNXPGY4TOUWELGB7/action/replication_record"}},"created_at":"2026-06-02T01:04:09.500194+00:00","updated_at":"2026-06-02T01:04:09.500194+00:00"}