{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:D2653UJ5LMDJHWHKNDXHAXN2CG","short_pith_number":"pith:D2653UJ5","schema_version":"1.0","canonical_sha256":"1ebdddd13d5b0693d8ea68ee705dba11b87f86f84b3565554cfd857e6f9f0af7","source":{"kind":"arxiv","id":"2608.04336","version":1},"attestation_state":"computed","paper":{"title":"COMPAS: Difficulty-Aware Joint Search for Optimizing Code Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Dong Huang, Gunel Jahangirova, Jie M. Zhang, Jingzhi Gong, Mark Harman, Mohammad Reza Mousavi","submitted_at":"2026-08-05T01:28:43Z","abstract_excerpt":"Code generation systems make each LLM call with a model, a prompt, and decoding settings. However, existing optimization methods usually tune only part of these choices or use one fixed configuration for all tasks: global optimizers search one configuration for all tasks, routers choose only a model, and prompt optimizers keep the model and decoding settings fixed. This leaves their joint, group-specific interactions unclear. We therefore examine how these choices interact and observe that prompts and decoding settings interact, tuning effects vary by model, and the best configuration varies b"},"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":"2608.04336","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-08-05T01:28:43Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"90637ed2d312983a00b0a1a01fb96b8b047ba6037dec9774f46c461223123d99","abstract_canon_sha256":"cd083a90759d46947aaef33c2cac78da1b66bd67e0f99f1343a0372f1659ab22"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:33:01.085330Z","signature_b64":"PisUnWYI+oZ3ExmXqnsfWywk7aajeYzxgBvfPzRn4VTzPKEdZO6VxJ+gb9ujGvGkqzRO/AzoRdecSQOiP75KAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1ebdddd13d5b0693d8ea68ee705dba11b87f86f84b3565554cfd857e6f9f0af7","last_reissued_at":"2026-08-06T01:33:01.083815Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:33:01.083815Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"COMPAS: Difficulty-Aware Joint Search for Optimizing Code Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Dong Huang, Gunel Jahangirova, Jie M. Zhang, Jingzhi Gong, Mark Harman, Mohammad Reza Mousavi","submitted_at":"2026-08-05T01:28:43Z","abstract_excerpt":"Code generation systems make each LLM call with a model, a prompt, and decoding settings. However, existing optimization methods usually tune only part of these choices or use one fixed configuration for all tasks: global optimizers search one configuration for all tasks, routers choose only a model, and prompt optimizers keep the model and decoding settings fixed. This leaves their joint, group-specific interactions unclear. We therefore examine how these choices interact and observe that prompts and decoding settings interact, tuning effects vary by model, and the best configuration varies b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04336","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/2608.04336/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":"2608.04336","created_at":"2026-08-06T01:33:01.085683+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.04336v1","created_at":"2026-08-06T01:33:01.085683+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04336","created_at":"2026-08-06T01:33:01.085683+00:00"},{"alias_kind":"pith_short_12","alias_value":"D2653UJ5LMDJ","created_at":"2026-08-06T01:33:01.085683+00:00"},{"alias_kind":"pith_short_16","alias_value":"D2653UJ5LMDJHWHK","created_at":"2026-08-06T01:33:01.085683+00:00"},{"alias_kind":"pith_short_8","alias_value":"D2653UJ5","created_at":"2026-08-06T01:33:01.085683+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/D2653UJ5LMDJHWHKNDXHAXN2CG","json":"https://pith.science/pith/D2653UJ5LMDJHWHKNDXHAXN2CG.json","graph_json":"https://pith.science/api/pith-number/D2653UJ5LMDJHWHKNDXHAXN2CG/graph.json","events_json":"https://pith.science/api/pith-number/D2653UJ5LMDJHWHKNDXHAXN2CG/events.json","paper":"https://pith.science/paper/D2653UJ5"},"agent_actions":{"view_html":"https://pith.science/pith/D2653UJ5LMDJHWHKNDXHAXN2CG","download_json":"https://pith.science/pith/D2653UJ5LMDJHWHKNDXHAXN2CG.json","view_paper":"https://pith.science/paper/D2653UJ5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.04336&json=true","fetch_graph":"https://pith.science/api/pith-number/D2653UJ5LMDJHWHKNDXHAXN2CG/graph.json","fetch_events":"https://pith.science/api/pith-number/D2653UJ5LMDJHWHKNDXHAXN2CG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/D2653UJ5LMDJHWHKNDXHAXN2CG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/D2653UJ5LMDJHWHKNDXHAXN2CG/action/storage_attestation","attest_author":"https://pith.science/pith/D2653UJ5LMDJHWHKNDXHAXN2CG/action/author_attestation","sign_citation":"https://pith.science/pith/D2653UJ5LMDJHWHKNDXHAXN2CG/action/citation_signature","submit_replication":"https://pith.science/pith/D2653UJ5LMDJHWHKNDXHAXN2CG/action/replication_record"}},"created_at":"2026-08-06T01:33:01.085683+00:00","updated_at":"2026-08-06T01:33:01.085683+00:00"}