{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YY44VUGWUAS22WPVYNBP7M5PQQ","short_pith_number":"pith:YY44VUGW","schema_version":"1.0","canonical_sha256":"c639cad0d6a025ad59f5c342ffb3af843bff412ca3fea0d257048fc19a35c60f","source":{"kind":"arxiv","id":"2409.12993","version":2},"attestation_state":"computed","paper":{"title":"CraftRTL: High-quality Synthetic Data Generation for Verilog Code Models with Correct-by-Construction Non-Textual Representations and Targeted Code Repair","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AR","authors_text":"Haoxing Ren, Mingjie Liu, Wenfei Zhou, Yun-Da Tsai","submitted_at":"2024-09-19T12:15:55Z","abstract_excerpt":"Despite the significant progress made in code generation with large language models, challenges persist, especially with hardware description languages such as Verilog. This paper first presents an analysis of fine-tuned LLMs on Verilog coding, with synthetic data from prior methods. We identify two main issues: difficulties in handling non-textual representations (Karnaugh maps, state-transition diagrams and waveforms) and significant variability during training with models randomly making \"minor\" mistakes. To address these limitations, we enhance data curation by creating correct-by-construc"},"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":"2409.12993","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AR","submitted_at":"2024-09-19T12:15:55Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"51f971200bb5fbf992fc7ed79c9cde108093df6845bb3024ff42aaf7055ade68","abstract_canon_sha256":"fc7f5b29fd35b7a0faebe017ffd0a7677d7c15d4a19a50720384ad5ba560fcde"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:10:44.053358Z","signature_b64":"ykf0EQSRoegP4JGMh8oVo09Bk4UykafAABHTrtPjYmoW43toa51o7/M8hjWn+CZlqBzeqtlsN3JmuayxiGYKCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c639cad0d6a025ad59f5c342ffb3af843bff412ca3fea0d257048fc19a35c60f","last_reissued_at":"2026-07-05T10:10:44.052918Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:10:44.052918Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CraftRTL: High-quality Synthetic Data Generation for Verilog Code Models with Correct-by-Construction Non-Textual Representations and Targeted Code Repair","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AR","authors_text":"Haoxing Ren, Mingjie Liu, Wenfei Zhou, Yun-Da Tsai","submitted_at":"2024-09-19T12:15:55Z","abstract_excerpt":"Despite the significant progress made in code generation with large language models, challenges persist, especially with hardware description languages such as Verilog. This paper first presents an analysis of fine-tuned LLMs on Verilog coding, with synthetic data from prior methods. We identify two main issues: difficulties in handling non-textual representations (Karnaugh maps, state-transition diagrams and waveforms) and significant variability during training with models randomly making \"minor\" mistakes. To address these limitations, we enhance data curation by creating correct-by-construc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.12993","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/2409.12993/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":"2409.12993","created_at":"2026-07-05T10:10:44.052975+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.12993v2","created_at":"2026-07-05T10:10:44.052975+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.12993","created_at":"2026-07-05T10:10:44.052975+00:00"},{"alias_kind":"pith_short_12","alias_value":"YY44VUGWUAS2","created_at":"2026-07-05T10:10:44.052975+00:00"},{"alias_kind":"pith_short_16","alias_value":"YY44VUGWUAS22WPV","created_at":"2026-07-05T10:10:44.052975+00:00"},{"alias_kind":"pith_short_8","alias_value":"YY44VUGW","created_at":"2026-07-05T10:10:44.052975+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23759","citing_title":"VeriPilot: An LLM-Powered Verilog Debugging Framework","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05680","citing_title":"CASS-RTL: Correctness-Aware Subspace Steering for RTL Generation with LLMs","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28279","citing_title":"Agentic Hardware Design as Repository-Level Code Evolution","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2507.04736","citing_title":"ChipSeek: Optimizing Verilog Generation via EDA-Integrated Reinforcement Learning","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23602","citing_title":"TimingLLM: A Two-Stage Retrieval-Augmented Framework for Pre-Synthesis Timing Prediction from Verilog","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ","json":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ.json","graph_json":"https://pith.science/api/pith-number/YY44VUGWUAS22WPVYNBP7M5PQQ/graph.json","events_json":"https://pith.science/api/pith-number/YY44VUGWUAS22WPVYNBP7M5PQQ/events.json","paper":"https://pith.science/paper/YY44VUGW"},"agent_actions":{"view_html":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ","download_json":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ.json","view_paper":"https://pith.science/paper/YY44VUGW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.12993&json=true","fetch_graph":"https://pith.science/api/pith-number/YY44VUGWUAS22WPVYNBP7M5PQQ/graph.json","fetch_events":"https://pith.science/api/pith-number/YY44VUGWUAS22WPVYNBP7M5PQQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ/action/storage_attestation","attest_author":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ/action/author_attestation","sign_citation":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ/action/citation_signature","submit_replication":"https://pith.science/pith/YY44VUGWUAS22WPVYNBP7M5PQQ/action/replication_record"}},"created_at":"2026-07-05T10:10:44.052975+00:00","updated_at":"2026-07-05T10:10:44.052975+00:00"}