{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2DPUOPBUTSOVJKPII57CYKQQPT","short_pith_number":"pith:2DPUOPBU","schema_version":"1.0","canonical_sha256":"d0df473c349c9d54a9e8477e2c2a107ce007f73e8b88bb11a81dbd27895f7df3","source":{"kind":"arxiv","id":"2502.05111","version":2},"attestation_state":"computed","paper":{"title":"Flexible and Efficient Grammar-Constrained Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Kanghee Park, Loris D'Antoni, Timothy Zhou","submitted_at":"2025-02-07T17:35:17Z","abstract_excerpt":"Large Language Models (LLMs) are often asked to generate structured outputs that obey precise syntactic rules, such as code snippets or formatted data. Grammar-constrained decoding (GCD) can guarantee that LLM outputs matches such rules by masking out tokens that will provably lead to outputs that do not belong to a specified context-free grammar (CFG). To guarantee soundness, GCD algorithms have to compute how a given LLM subword tokenizer can align with the tokens used\n  by a given context-free grammar and compute token masks based on this information. Doing so efficiently is challenging and"},"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":"2502.05111","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-07T17:35:17Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0a3f8c5f86c73b49c409f94bb3567f21b1fe78ea6b71d34734a7ddc4a05996f2","abstract_canon_sha256":"096c1760740924c123c54d76e19ed02cc45edc978c9f003f18fa0f89f6894a20"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:43.271665Z","signature_b64":"FirHDQH+c9E1/EinL4NLHAhyb65ACYJ/YYABVw9i4bfT1dmN79m5KHaMYdfqdnALxJNZUCa99O5ir1flvOCFCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d0df473c349c9d54a9e8477e2c2a107ce007f73e8b88bb11a81dbd27895f7df3","last_reissued_at":"2026-07-05T11:37:43.271078Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:43.271078Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Flexible and Efficient Grammar-Constrained Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Kanghee Park, Loris D'Antoni, Timothy Zhou","submitted_at":"2025-02-07T17:35:17Z","abstract_excerpt":"Large Language Models (LLMs) are often asked to generate structured outputs that obey precise syntactic rules, such as code snippets or formatted data. Grammar-constrained decoding (GCD) can guarantee that LLM outputs matches such rules by masking out tokens that will provably lead to outputs that do not belong to a specified context-free grammar (CFG). To guarantee soundness, GCD algorithms have to compute how a given LLM subword tokenizer can align with the tokens used\n  by a given context-free grammar and compute token masks based on this information. Doing so efficiently is challenging and"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.05111","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/2502.05111/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":"2502.05111","created_at":"2026-07-05T11:37:43.271146+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.05111v2","created_at":"2026-07-05T11:37:43.271146+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.05111","created_at":"2026-07-05T11:37:43.271146+00:00"},{"alias_kind":"pith_short_12","alias_value":"2DPUOPBUTSOV","created_at":"2026-07-05T11:37:43.271146+00:00"},{"alias_kind":"pith_short_16","alias_value":"2DPUOPBUTSOVJKPI","created_at":"2026-07-05T11:37:43.271146+00:00"},{"alias_kind":"pith_short_8","alias_value":"2DPUOPBU","created_at":"2026-07-05T11:37:43.271146+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.01926","citing_title":"Mitigating Bias in Locally Constrained Decoding via Tractable Proposals","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31808","citing_title":"Large Databases Need Small, Open-Weight Language Models","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2510.14703","citing_title":"ToolPRM: Fine-Grained Inference Scaling of Structured Outputs for Function Calling","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10065","citing_title":"NCO: A Versatile Plug-in for Handling Negative Constraints in Decoding","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14862","citing_title":"Schema Key Wording as an Instruction Channel in Structured Generation under Constrained Decoding","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT","json":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT.json","graph_json":"https://pith.science/api/pith-number/2DPUOPBUTSOVJKPII57CYKQQPT/graph.json","events_json":"https://pith.science/api/pith-number/2DPUOPBUTSOVJKPII57CYKQQPT/events.json","paper":"https://pith.science/paper/2DPUOPBU"},"agent_actions":{"view_html":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT","download_json":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT.json","view_paper":"https://pith.science/paper/2DPUOPBU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.05111&json=true","fetch_graph":"https://pith.science/api/pith-number/2DPUOPBUTSOVJKPII57CYKQQPT/graph.json","fetch_events":"https://pith.science/api/pith-number/2DPUOPBUTSOVJKPII57CYKQQPT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT/action/storage_attestation","attest_author":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT/action/author_attestation","sign_citation":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT/action/citation_signature","submit_replication":"https://pith.science/pith/2DPUOPBUTSOVJKPII57CYKQQPT/action/replication_record"}},"created_at":"2026-07-05T11:37:43.271146+00:00","updated_at":"2026-07-05T11:37:43.271146+00:00"}