{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:FCPQML7QH5IJCFDCRSOGCJZJJG","short_pith_number":"pith:FCPQML7Q","schema_version":"1.0","canonical_sha256":"289f062ff03f509114628c9c61272949b04e744a56208418a6b76a00608f050b","source":{"kind":"arxiv","id":"2504.04030","version":2},"attestation_state":"computed","paper":{"title":"OpenCodeInstruct: A Large-scale Instruction Tuning Dataset for Code LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.SE","authors_text":"Aleksander Ficek, Boris Ginsburg, Jocelyn Huang, Mehrzad Samadi, Somshubra Majumdar, Vahid Noroozi, Wasi Uddin Ahmad","submitted_at":"2025-04-05T02:52:16Z","abstract_excerpt":"Large Language Models (LLMs) have transformed software development by enabling code generation, automated debugging, and complex reasoning. However, their continued advancement is constrained by the scarcity of high-quality, publicly available supervised fine-tuning (SFT) datasets tailored for coding tasks. To bridge this gap, we introduce OpenCodeInstruct, the largest open-access instruction tuning dataset, comprising 5 million diverse samples. Each sample includes a programming question, solution, test cases, execution feedback, and LLM-generated quality assessments. We fine-tune various bas"},"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":"2504.04030","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2025-04-05T02:52:16Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"643e87b4465c9d6842aefc4ac08b1d595255a90d538d2005405eff6cefde37ba","abstract_canon_sha256":"2b74c6abd7ad6bf75ed9839a04d76eb7072be951819423c01ac7d731adff2973"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:50:34.499161Z","signature_b64":"MZFeaUSG24gaPijAcunfPNqR94dzoB0Dg9UWHTeHsB0mG4+qAbj8Z63wMrFWxPlF5C/5ZGP2NXyNnEQ4tuhsCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"289f062ff03f509114628c9c61272949b04e744a56208418a6b76a00608f050b","last_reissued_at":"2026-07-05T11:50:34.498710Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:50:34.498710Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OpenCodeInstruct: A Large-scale Instruction Tuning Dataset for Code LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.SE","authors_text":"Aleksander Ficek, Boris Ginsburg, Jocelyn Huang, Mehrzad Samadi, Somshubra Majumdar, Vahid Noroozi, Wasi Uddin Ahmad","submitted_at":"2025-04-05T02:52:16Z","abstract_excerpt":"Large Language Models (LLMs) have transformed software development by enabling code generation, automated debugging, and complex reasoning. However, their continued advancement is constrained by the scarcity of high-quality, publicly available supervised fine-tuning (SFT) datasets tailored for coding tasks. To bridge this gap, we introduce OpenCodeInstruct, the largest open-access instruction tuning dataset, comprising 5 million diverse samples. Each sample includes a programming question, solution, test cases, execution feedback, and LLM-generated quality assessments. We fine-tune various bas"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.04030","kind":"arxiv","version":2},"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/2504.04030/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":"2504.04030","created_at":"2026-07-05T11:50:34.498767+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.04030v2","created_at":"2026-07-05T11:50:34.498767+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.04030","created_at":"2026-07-05T11:50:34.498767+00:00"},{"alias_kind":"pith_short_12","alias_value":"FCPQML7QH5IJ","created_at":"2026-07-05T11:50:34.498767+00:00"},{"alias_kind":"pith_short_16","alias_value":"FCPQML7QH5IJCFDC","created_at":"2026-07-05T11:50:34.498767+00:00"},{"alias_kind":"pith_short_8","alias_value":"FCPQML7Q","created_at":"2026-07-05T11:50:34.498767+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.18327","citing_title":"Self-CTRL: Self-Consistency Training with Reinforcement Learning","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11817","citing_title":"Grammar-Constrained Decoding Can Jailbreak LLMs into Generating Malicious Code","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2606.18286","citing_title":"CODEBLOCK: Learning to Supervise Code at the Right Granularity","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09145","citing_title":"PrivCode++: Latent-Conditioned Differentially Private Code Generation for Comprehensive Guarantees","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27617","citing_title":"Masked Language Flow Models","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28998","citing_title":"Reward-Free Code Alignment from Pretrained or Fine-Tuned LLM: Unpacking the Trade-offs for Code Generation","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25296","citing_title":"Subjective Code Preferences in Experts and Large Language Models","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07711","citing_title":"SimCT: Recovering Lost Supervision for Cross-Tokenizer On-Policy Distillation","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08302","citing_title":"DMax: Aggressive Parallel Decoding for dLLMs","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2512.20856","citing_title":"NVIDIA Nemotron 3: Efficient and Open Intelligence","ref_index":146,"is_internal_anchor":false},{"citing_arxiv_id":"2602.08813","citing_title":"Robust Policy Optimization to Prevent Catastrophic Forgetting","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.00824","citing_title":"Yet Even Less Is Even Better For Agentic, Reasoning, and Coding LLMs","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08302","citing_title":"DMax: Aggressive Parallel Decoding for dLLMs","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07711","citing_title":"SimCT: Recovering Lost Supervision for Cross-Tokenizer On-Policy Distillation","ref_index":52,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG","json":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG.json","graph_json":"https://pith.science/api/pith-number/FCPQML7QH5IJCFDCRSOGCJZJJG/graph.json","events_json":"https://pith.science/api/pith-number/FCPQML7QH5IJCFDCRSOGCJZJJG/events.json","paper":"https://pith.science/paper/FCPQML7Q"},"agent_actions":{"view_html":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG","download_json":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG.json","view_paper":"https://pith.science/paper/FCPQML7Q","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.04030&json=true","fetch_graph":"https://pith.science/api/pith-number/FCPQML7QH5IJCFDCRSOGCJZJJG/graph.json","fetch_events":"https://pith.science/api/pith-number/FCPQML7QH5IJCFDCRSOGCJZJJG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG/action/storage_attestation","attest_author":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG/action/author_attestation","sign_citation":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG/action/citation_signature","submit_replication":"https://pith.science/pith/FCPQML7QH5IJCFDCRSOGCJZJJG/action/replication_record"}},"created_at":"2026-07-05T11:50:34.498767+00:00","updated_at":"2026-07-05T11:50:34.498767+00:00"}