{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:V325N4CRRLIT5APD3CBBVITJG7","short_pith_number":"pith:V325N4CR","schema_version":"1.0","canonical_sha256":"aef5d6f0518ad13e81e3d8821aa26937e174a913ea3c3173c7be1fa9fc5384c0","source":{"kind":"arxiv","id":"2207.01780","version":3},"attestation_state":"computed","paper":{"title":"CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.PL"],"primary_cat":"cs.LG","authors_text":"Akhilesh Deepak Gotmare, Hung Le, Silvio Savarese, Steven C.H. Hoi, Yue Wang","submitted_at":"2022-07-05T02:42:15Z","abstract_excerpt":"Program synthesis or code generation aims to generate a program that satisfies a problem specification. Recent approaches using large-scale pretrained language models (LMs) have shown promising results, yet they have some critical limitations. In particular, they often follow a standard supervised fine-tuning procedure to train a code generation model only from the pairs of natural-language problem descriptions and ground-truth programs. Such paradigm largely ignores some important but potentially useful signals in the problem specification such as unit tests, which thus often results in poor "},"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":"2207.01780","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-07-05T02:42:15Z","cross_cats_sorted":["cs.CL","cs.PL"],"title_canon_sha256":"a80c9aeb8cf3af35533f573adcb781e09f10cb093234b239e9648f6fe2fffdbf","abstract_canon_sha256":"627310f584db4f7fcc3101dad68a69b184ba4d2a28d1fc0e8b6ec140be0a15e2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:12:52.214261Z","signature_b64":"2P1NPsb9o62Q1D7Tg4O48fQ7pvFWFKQ4HKmaO2rtrF/Eae/6Wo6UkTM9JOKTiKsEOWJKAJBuRDiam9AuR9lODA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aef5d6f0518ad13e81e3d8821aa26937e174a913ea3c3173c7be1fa9fc5384c0","last_reissued_at":"2026-07-05T05:12:52.213784Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:12:52.213784Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.PL"],"primary_cat":"cs.LG","authors_text":"Akhilesh Deepak Gotmare, Hung Le, Silvio Savarese, Steven C.H. Hoi, Yue Wang","submitted_at":"2022-07-05T02:42:15Z","abstract_excerpt":"Program synthesis or code generation aims to generate a program that satisfies a problem specification. Recent approaches using large-scale pretrained language models (LMs) have shown promising results, yet they have some critical limitations. In particular, they often follow a standard supervised fine-tuning procedure to train a code generation model only from the pairs of natural-language problem descriptions and ground-truth programs. Such paradigm largely ignores some important but potentially useful signals in the problem specification such as unit tests, which thus often results in poor "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.01780","kind":"arxiv","version":3},"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/2207.01780/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":"2207.01780","created_at":"2026-07-05T05:12:52.213842+00:00"},{"alias_kind":"arxiv_version","alias_value":"2207.01780v3","created_at":"2026-07-05T05:12:52.213842+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.01780","created_at":"2026-07-05T05:12:52.213842+00:00"},{"alias_kind":"pith_short_12","alias_value":"V325N4CRRLIT","created_at":"2026-07-05T05:12:52.213842+00:00"},{"alias_kind":"pith_short_16","alias_value":"V325N4CRRLIT5APD","created_at":"2026-07-05T05:12:52.213842+00:00"},{"alias_kind":"pith_short_8","alias_value":"V325N4CR","created_at":"2026-07-05T05:12:52.213842+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":15,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.05253","citing_title":"Alpha-RTL: Test-Time Training for RTL Hardware Optimization","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24375","citing_title":"Distilling Game Code World Model Generation into Lightweight Large Language Models","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26807","citing_title":"HTMLCure: Turning Browser Experience into State Guided Repair for Interactive HTML","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30478","citing_title":"Improving Small Language Models for Code Generation with Reinforcement Learning from Verification Feedback","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28409","citing_title":"Efficient Post-training of LLMs for Code Generation With Offline Reinforcement Learning","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2509.26383","citing_title":"Efficient and Transferable Agentic Knowledge Graph RAG via Reinforcement Learning","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2510.16416","citing_title":"SSL4RL: Revisiting Self-supervised Learning as Intrinsic Reward for Visual-Language Reasoning","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2207.10397","citing_title":"CodeT: Code Generation with Generated Tests","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13935","citing_title":"Beyond Mode-Seeking RL: Trajectory-Balance Post-Training for Diffusion Language Models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11680","citing_title":"ShapeCodeBench: A Renewable Benchmark for Perception-to-Program Reconstruction of Synthetic Shape Scenes","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08468","citing_title":"PYTHALAB-MERA: Validation-Grounded Memory, Retrieval, and Acceptance Control for Frozen-LLM Coding Agents","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02913","citing_title":"Generate, Filter, Control, Replay: A Comprehensive Survey of Rollout Strategies for LLM Reinforcement Learning","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2303.17651","citing_title":"Self-Refine: Iterative Refinement with Self-Feedback","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2305.16291","citing_title":"Voyager: An Open-Ended Embodied Agent with Large Language Models","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16804","citing_title":"AutoOR: Scalably Post-training LLMs to Autoformalize Operations Research Problems","ref_index":44,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7","json":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7.json","graph_json":"https://pith.science/api/pith-number/V325N4CRRLIT5APD3CBBVITJG7/graph.json","events_json":"https://pith.science/api/pith-number/V325N4CRRLIT5APD3CBBVITJG7/events.json","paper":"https://pith.science/paper/V325N4CR"},"agent_actions":{"view_html":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7","download_json":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7.json","view_paper":"https://pith.science/paper/V325N4CR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2207.01780&json=true","fetch_graph":"https://pith.science/api/pith-number/V325N4CRRLIT5APD3CBBVITJG7/graph.json","fetch_events":"https://pith.science/api/pith-number/V325N4CRRLIT5APD3CBBVITJG7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7/action/storage_attestation","attest_author":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7/action/author_attestation","sign_citation":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7/action/citation_signature","submit_replication":"https://pith.science/pith/V325N4CRRLIT5APD3CBBVITJG7/action/replication_record"}},"created_at":"2026-07-05T05:12:52.213842+00:00","updated_at":"2026-07-05T05:12:52.213842+00:00"}