{"paper":{"title":"FlowPlan-G2P: A Structured Generation Framework for Transforming Scientific Papers into Patent Descriptions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A structured graph framework turns scientific papers into patent descriptions more effectively than scaling up language models.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Kris W Pan, Yongmin Yoo","submitted_at":"2026-01-05T22:40:15Z","abstract_excerpt":"Generating patent descriptions from scientific papers is challenging due to fundamental rhetorical and structural disparities between the two genres. Existing approaches treat this as surface-level rewriting, failing to capture the hierarchical reasoning and statutory constraints inherent in patent drafting. We propose FlowPlan-G2P, a graph-mediated generation framework that decomposes this transformation into three stages: (1) Concept Graph Induction, extracting technical entities and functional dependencies into a directed graph; (2) Section-level Planning, partitioning the graph into cohere"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"structured decomposition is a stronger determinant of quality than model scale","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the expert-validated benchmarks and domain-specific evaluation accurately measure legal compliance and that the induced concept graphs capture all statutory constraints needed for valid patent descriptions","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"FlowPlan-G2P decomposes scientific paper to patent conversion into concept graph induction, section-level planning, and graph-conditioned generation, outperforming direct proprietary models under a domain-specific legal compliance evaluation.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A structured graph framework turns scientific papers into patent descriptions more effectively than scaling up language models.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7f466ad62415d41612dbbdaabcbfe203d8b53895743523518c0465a61a87614f"},"source":{"id":"2601.02589","kind":"arxiv","version":4},"verdict":{"id":"3e75c7dd-7a45-4121-b58f-3512ff3df041","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T17:17:53.863280Z","strongest_claim":"structured decomposition is a stronger determinant of quality than model scale","one_line_summary":"FlowPlan-G2P decomposes scientific paper to patent conversion into concept graph induction, section-level planning, and graph-conditioned generation, outperforming direct proprietary models under a domain-specific legal compliance evaluation.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the expert-validated benchmarks and domain-specific evaluation accurately measure legal compliance and that the induced concept graphs capture all statutory constraints needed for valid patent descriptions","pith_extraction_headline":"A structured graph framework turns scientific papers into patent descriptions more effectively than scaling up language models."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2601.02589/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":1,"snapshot_sha256":"ab1a771bd7d4c8ef841ff6222bb1c858cfbbfb04240a476945216f0c2ed7a7eb"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}