{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UUOM6PQD4E6RGE6SIH4LJKTVYY","short_pith_number":"pith:UUOM6PQD","schema_version":"1.0","canonical_sha256":"a51ccf3e03e13d1313d241f8b4aa75c62e4f84f83b888180675767d8958c6d89","source":{"kind":"arxiv","id":"2506.09396","version":1},"attestation_state":"computed","paper":{"title":"Reasoning as a Resource: Optimizing Fast and Slow Thinking in Code Generation Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Shuai Wang, Zongjie Li","submitted_at":"2025-06-11T04:55:00Z","abstract_excerpt":"This position paper proposes a fundamental shift in designing code generation models: treating reasoning depth as a controllable resource. Rather than being an incidental byproduct of prompting, we argue that the trade-off between rapid, direct answers (\"fast thinking\") and elaborate, chain-of-thought deliberation (\"slow thinking\") must be explicitly managed. We contend that optimizing reasoning budgets across the entire model lifecycle - from synthetic data creation and benchmarking to real-world deploymen - can unlock superior trade-offs among accuracy, latency, and cost. This paper outlines"},"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":"2506.09396","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2025-06-11T04:55:00Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"74c82050b459d19cac15bd5ebece77935279672460b288381dd49d8d2166f72b","abstract_canon_sha256":"b54faebf0c1e25033b07774d8a03c990686d0b1fb44a93101bb8fb1947732369"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:19:36.371309Z","signature_b64":"9SQoh86MmpJjQbCosw6jSCBh2AknvONhX9pvTmojODq/087LnDN1Y6zxA2kasWYw2EW34d4bqagcjksEcN9CDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a51ccf3e03e13d1313d241f8b4aa75c62e4f84f83b888180675767d8958c6d89","last_reissued_at":"2026-07-05T11:19:36.370773Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:19:36.370773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reasoning as a Resource: Optimizing Fast and Slow Thinking in Code Generation Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Shuai Wang, Zongjie Li","submitted_at":"2025-06-11T04:55:00Z","abstract_excerpt":"This position paper proposes a fundamental shift in designing code generation models: treating reasoning depth as a controllable resource. Rather than being an incidental byproduct of prompting, we argue that the trade-off between rapid, direct answers (\"fast thinking\") and elaborate, chain-of-thought deliberation (\"slow thinking\") must be explicitly managed. We contend that optimizing reasoning budgets across the entire model lifecycle - from synthetic data creation and benchmarking to real-world deploymen - can unlock superior trade-offs among accuracy, latency, and cost. This paper outlines"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.09396","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/2506.09396/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":"2506.09396","created_at":"2026-07-05T11:19:36.370832+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.09396v1","created_at":"2026-07-05T11:19:36.370832+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.09396","created_at":"2026-07-05T11:19:36.370832+00:00"},{"alias_kind":"pith_short_12","alias_value":"UUOM6PQD4E6R","created_at":"2026-07-05T11:19:36.370832+00:00"},{"alias_kind":"pith_short_16","alias_value":"UUOM6PQD4E6RGE6S","created_at":"2026-07-05T11:19:36.370832+00:00"},{"alias_kind":"pith_short_8","alias_value":"UUOM6PQD","created_at":"2026-07-05T11:19:36.370832+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.09135","citing_title":"Position Paper: Programming Language Techniques for Bridging LLM Code Generation Semantic Gaps","ref_index":30,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY","json":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY.json","graph_json":"https://pith.science/api/pith-number/UUOM6PQD4E6RGE6SIH4LJKTVYY/graph.json","events_json":"https://pith.science/api/pith-number/UUOM6PQD4E6RGE6SIH4LJKTVYY/events.json","paper":"https://pith.science/paper/UUOM6PQD"},"agent_actions":{"view_html":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY","download_json":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY.json","view_paper":"https://pith.science/paper/UUOM6PQD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.09396&json=true","fetch_graph":"https://pith.science/api/pith-number/UUOM6PQD4E6RGE6SIH4LJKTVYY/graph.json","fetch_events":"https://pith.science/api/pith-number/UUOM6PQD4E6RGE6SIH4LJKTVYY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY/action/storage_attestation","attest_author":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY/action/author_attestation","sign_citation":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY/action/citation_signature","submit_replication":"https://pith.science/pith/UUOM6PQD4E6RGE6SIH4LJKTVYY/action/replication_record"}},"created_at":"2026-07-05T11:19:36.370832+00:00","updated_at":"2026-07-05T11:19:36.370832+00:00"}