{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XIHZVD46O5Y5226XBUSHWPVSNZ","short_pith_number":"pith:XIHZVD46","schema_version":"1.0","canonical_sha256":"ba0f9a8f9e7771dd6bd70d247b3eb26e676940ac2197e274f4836ca45fe7ea09","source":{"kind":"arxiv","id":"2502.01718","version":4},"attestation_state":"computed","paper":{"title":"ACECODER: Acing Coder RL via Automated Test-Case Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.SE","authors_text":"Dongfu Jiang, Haozhe Wang, Huaye Zeng, Ping Nie, Wenhu Chen, Xiaotong Chen","submitted_at":"2025-02-03T18:46:04Z","abstract_excerpt":"Most progress in recent coder models has been driven by supervised fine-tuning (SFT), while the potential of reinforcement learning (RL) remains largely unexplored, primarily due to the lack of reliable reward data/model in the code domain. In this paper, we address this challenge by leveraging automated large-scale test-case synthesis to enhance code model training. Specifically, we design a pipeline that generates extensive (question, test-cases) pairs from existing code data. Using these test cases, we construct preference pairs based on pass rates over sampled programs to train reward mode"},"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":"2502.01718","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-02-03T18:46:04Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"817b6693d839f17264e6f5c933eebffe90eb9e37f36b9af1d59184e1cc356b9d","abstract_canon_sha256":"501d1168dc19a577811ef10844f8ee41ffee9adc20086fd98e18078fb265da70"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:08:39.972898Z","signature_b64":"LyyVqHO7a1XJn/EccyiVB1E294q9OH5dCxwDCSWZzzQbKYP9A6XmVPXSW0+3JMTmbFT+nWktizJnopI6vVuKBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba0f9a8f9e7771dd6bd70d247b3eb26e676940ac2197e274f4836ca45fe7ea09","last_reissued_at":"2026-07-05T11:08:39.972411Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:08:39.972411Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ACECODER: Acing Coder RL via Automated Test-Case Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.SE","authors_text":"Dongfu Jiang, Haozhe Wang, Huaye Zeng, Ping Nie, Wenhu Chen, Xiaotong Chen","submitted_at":"2025-02-03T18:46:04Z","abstract_excerpt":"Most progress in recent coder models has been driven by supervised fine-tuning (SFT), while the potential of reinforcement learning (RL) remains largely unexplored, primarily due to the lack of reliable reward data/model in the code domain. In this paper, we address this challenge by leveraging automated large-scale test-case synthesis to enhance code model training. Specifically, we design a pipeline that generates extensive (question, test-cases) pairs from existing code data. Using these test cases, we construct preference pairs based on pass rates over sampled programs to train reward mode"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.01718","kind":"arxiv","version":4},"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/2502.01718/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":"2502.01718","created_at":"2026-07-05T11:08:39.972473+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.01718v4","created_at":"2026-07-05T11:08:39.972473+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.01718","created_at":"2026-07-05T11:08:39.972473+00:00"},{"alias_kind":"pith_short_12","alias_value":"XIHZVD46O5Y5","created_at":"2026-07-05T11:08:39.972473+00:00"},{"alias_kind":"pith_short_16","alias_value":"XIHZVD46O5Y5226X","created_at":"2026-07-05T11:08:39.972473+00:00"},{"alias_kind":"pith_short_8","alias_value":"XIHZVD46","created_at":"2026-07-05T11:08:39.972473+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.29398","citing_title":"GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2502.18449","citing_title":"SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07864","citing_title":"ZeroCoder: Can LLMs Improve Code Generation Without Ground-Truth Supervision?","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18473","citing_title":"Train Separately, Merge Together: Modular Post-Training with Mixture-of-Experts","ref_index":48,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ","json":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ.json","graph_json":"https://pith.science/api/pith-number/XIHZVD46O5Y5226XBUSHWPVSNZ/graph.json","events_json":"https://pith.science/api/pith-number/XIHZVD46O5Y5226XBUSHWPVSNZ/events.json","paper":"https://pith.science/paper/XIHZVD46"},"agent_actions":{"view_html":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ","download_json":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ.json","view_paper":"https://pith.science/paper/XIHZVD46","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.01718&json=true","fetch_graph":"https://pith.science/api/pith-number/XIHZVD46O5Y5226XBUSHWPVSNZ/graph.json","fetch_events":"https://pith.science/api/pith-number/XIHZVD46O5Y5226XBUSHWPVSNZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ/action/storage_attestation","attest_author":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ/action/author_attestation","sign_citation":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ/action/citation_signature","submit_replication":"https://pith.science/pith/XIHZVD46O5Y5226XBUSHWPVSNZ/action/replication_record"}},"created_at":"2026-07-05T11:08:39.972473+00:00","updated_at":"2026-07-05T11:08:39.972473+00:00"}