{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CM42U3X3FITAZYYQDZUBCNN4R3","short_pith_number":"pith:CM42U3X3","schema_version":"1.0","canonical_sha256":"1339aa6efb2a260ce3101e681135bc8ec59ba60014fafdfd1cada08a11c06915","source":{"kind":"arxiv","id":"2605.17174","version":1},"attestation_state":"computed","paper":{"title":"Beyond Execution: Static-Analysis Rewards and Hint-Conditioned Diffusion RL for Code Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Faroq AL-Tam, Jie M. Zhang, Muhammad Al-Qurishi, Shuyin Ouyang, Zhaozhi Qian","submitted_at":"2026-05-16T22:18:04Z","abstract_excerpt":"Reinforcement Learning (RL) is an important paradigm for aligning Diffusion Language Models (DLMs) toward functional correctness in code generation. However, these models often encounter a ``capability cliff'' on complex tasks, where execution-based semantic rewards become too low to provide a viable learning signal. In this paper, we present a systematic empirical study of RL post-training for diffusion-based code generation along three axes: reward design, hint-conditioned sampling, and task difficulty. We investigate the effectiveness of execution-free rewards as alternatives to traditional"},"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":"2605.17174","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SE","submitted_at":"2026-05-16T22:18:04Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f6ee007bbad4db077f1a9265ff09960797c0de5939e74b70691f071ee28070ac","abstract_canon_sha256":"653ab3c801622edcfa798f15995fe85961e69d3d4e49dd415389e07c78b8d088"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:03:43.380279Z","signature_b64":"iEfnBf/u/mvJ5yTMc5b71uBO8TSj9GSrhIqM/iNNCgZJsy1baLWjSkmo2xhy/y4+y/NBWUaS2ujOVcLXnGAjDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1339aa6efb2a260ce3101e681135bc8ec59ba60014fafdfd1cada08a11c06915","last_reissued_at":"2026-05-20T00:03:43.379502Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:03:43.379502Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Execution: Static-Analysis Rewards and Hint-Conditioned Diffusion RL for Code Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Faroq AL-Tam, Jie M. Zhang, Muhammad Al-Qurishi, Shuyin Ouyang, Zhaozhi Qian","submitted_at":"2026-05-16T22:18:04Z","abstract_excerpt":"Reinforcement Learning (RL) is an important paradigm for aligning Diffusion Language Models (DLMs) toward functional correctness in code generation. However, these models often encounter a ``capability cliff'' on complex tasks, where execution-based semantic rewards become too low to provide a viable learning signal. In this paper, we present a systematic empirical study of RL post-training for diffusion-based code generation along three axes: reward design, hint-conditioned sampling, and task difficulty. We investigate the effectiveness of execution-free rewards as alternatives to traditional"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17174","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/2605.17174/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-19T22:33:23.751160Z","status":"skipped","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T22:01:57.974772Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"b5cc19cd54d77ae96335ac2e55c2a75010d708f29bfad0ed72a756a37f0049b9"},"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":"2605.17174","created_at":"2026-05-20T00:03:43.379625+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.17174v1","created_at":"2026-05-20T00:03:43.379625+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17174","created_at":"2026-05-20T00:03:43.379625+00:00"},{"alias_kind":"pith_short_12","alias_value":"CM42U3X3FITA","created_at":"2026-05-20T00:03:43.379625+00:00"},{"alias_kind":"pith_short_16","alias_value":"CM42U3X3FITAZYYQ","created_at":"2026-05-20T00:03:43.379625+00:00"},{"alias_kind":"pith_short_8","alias_value":"CM42U3X3","created_at":"2026-05-20T00:03:43.379625+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/CM42U3X3FITAZYYQDZUBCNN4R3","json":"https://pith.science/pith/CM42U3X3FITAZYYQDZUBCNN4R3.json","graph_json":"https://pith.science/api/pith-number/CM42U3X3FITAZYYQDZUBCNN4R3/graph.json","events_json":"https://pith.science/api/pith-number/CM42U3X3FITAZYYQDZUBCNN4R3/events.json","paper":"https://pith.science/paper/CM42U3X3"},"agent_actions":{"view_html":"https://pith.science/pith/CM42U3X3FITAZYYQDZUBCNN4R3","download_json":"https://pith.science/pith/CM42U3X3FITAZYYQDZUBCNN4R3.json","view_paper":"https://pith.science/paper/CM42U3X3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.17174&json=true","fetch_graph":"https://pith.science/api/pith-number/CM42U3X3FITAZYYQDZUBCNN4R3/graph.json","fetch_events":"https://pith.science/api/pith-number/CM42U3X3FITAZYYQDZUBCNN4R3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CM42U3X3FITAZYYQDZUBCNN4R3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CM42U3X3FITAZYYQDZUBCNN4R3/action/storage_attestation","attest_author":"https://pith.science/pith/CM42U3X3FITAZYYQDZUBCNN4R3/action/author_attestation","sign_citation":"https://pith.science/pith/CM42U3X3FITAZYYQDZUBCNN4R3/action/citation_signature","submit_replication":"https://pith.science/pith/CM42U3X3FITAZYYQDZUBCNN4R3/action/replication_record"}},"created_at":"2026-05-20T00:03:43.379625+00:00","updated_at":"2026-05-20T00:03:43.379625+00:00"}