{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:HVR6JYBJ7NDJ5PFQG6V5U7EYMB","short_pith_number":"pith:HVR6JYBJ","schema_version":"1.0","canonical_sha256":"3d63e4e029fb469ebcb037abda7c986058a1e01fa154c3a0819a4e39cb49783f","source":{"kind":"arxiv","id":"2605.13793","version":1},"attestation_state":"computed","paper":{"title":"An LLM-Based System for Argument Reconstruction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"An LLM pipeline reconstructs natural language text into argument graphs with premises, conclusions, and support, attack or undercut relations.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Douglas Aldred, Fabio G. Cozman, Paulo Pirozelli, Victor Hugo Nascimento Rocha","submitted_at":"2026-05-13T17:13:45Z","abstract_excerpt":"Arguments are a fundamental aspect of human reasoning, in which claims are supported, challenged, and weighed against one another. We present an end-to-end large language model (LLM)-based system for reconstructing arguments from natural language text into abstract argument graphs. The system follows a multi-stage pipeline that progressively identifies argumentative components, selects relevant elements, and uncovers their logical relations. These elements are represented as directed acyclic graphs consisting of two component types (premises and conclusions) and three relation types (support, "},"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":true,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.13793","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-13T17:13:45Z","cross_cats_sorted":[],"title_canon_sha256":"343a4f7361632b6be5f355203c41c04410d921b181613586ab727426a3c397e8","abstract_canon_sha256":"8d0e79fb930445624dd1d57cbf1345c4fc9c5447549e226fd0ed062a77d84a30"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:15.596703Z","signature_b64":"wHPd3aVuD9a8uau7GfRW79t1mUaE/c9SQfrKRAApG8acW8OMmGaQQ9V5z8I8STvUVguG551JnGVLBeWwBycTAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3d63e4e029fb469ebcb037abda7c986058a1e01fa154c3a0819a4e39cb49783f","last_reissued_at":"2026-05-18T02:44:15.596283Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:15.596283Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"An LLM-Based System for Argument Reconstruction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"An LLM pipeline reconstructs natural language text into argument graphs with premises, conclusions, and support, attack or undercut relations.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Douglas Aldred, Fabio G. Cozman, Paulo Pirozelli, Victor Hugo Nascimento Rocha","submitted_at":"2026-05-13T17:13:45Z","abstract_excerpt":"Arguments are a fundamental aspect of human reasoning, in which claims are supported, challenged, and weighed against one another. We present an end-to-end large language model (LLM)-based system for reconstructing arguments from natural language text into abstract argument graphs. The system follows a multi-stage pipeline that progressively identifies argumentative components, selects relevant elements, and uncovers their logical relations. These elements are represented as directed acyclic graphs consisting of two component types (premises and conclusions) and three relation types (support, "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Results show that the system can adequately recover argumentative structures and, when adapted to different annotation schemes, achieve reasonable performance across benchmark datasets.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the multi-stage LLM pipeline reliably identifies argumentative components and their relations without systematic human correction or domain-specific fine-tuning beyond the described prompting.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"An LLM pipeline turns natural language arguments into structured graphs of premises, conclusions, and support/attack/undercut relations.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"An LLM pipeline reconstructs natural language text into argument graphs with premises, conclusions, and support, attack or undercut relations.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"68c3ea547e239f999238e80e9544e72d9224531778f855564cd54a676aba6247"},"source":{"id":"2605.13793","kind":"arxiv","version":1},"verdict":{"id":"6a086c8d-dfa4-492b-b91d-c5d694c40b83","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:07:42.537240Z","strongest_claim":"Results show that the system can adequately recover argumentative structures and, when adapted to different annotation schemes, achieve reasonable performance across benchmark datasets.","one_line_summary":"An LLM pipeline turns natural language arguments into structured graphs of premises, conclusions, and support/attack/undercut relations.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the multi-stage LLM pipeline reliably identifies argumentative components and their relations without systematic human correction or domain-specific fine-tuning beyond the described prompting.","pith_extraction_headline":"An LLM pipeline reconstructs natural language text into argument graphs with premises, conclusions, and support, attack or undercut relations."},"references":{"count":12,"sample":[{"doi":"","year":2010,"title":"Abstract dialectical frameworks.Proceedings of the Twelfth International Conference on Principles of Knowledge Representation and Reason- ing (KR 2010), pp","work_id":"ddf80b04-cebc-4b84-8d9e-e653647dbc99","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2005,"title":"Claudette Cayrol and Marie-Christine Lagasquie-Schiex. On the acceptability of arguments in bipo- lar argumentation frameworks.Proceedings of the Eighth European Conference on Symbolic and Quantitativ","work_id":"d8af82ca-5641-4976-a181-e26583a7e4a9","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Do emotions really affect argument convincingness? a dynamic approach with LLM-based manipulation checks","work_id":"bb06017d-6963-4a8b-9eff-f945073a09fe","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"ISBN 979-8-89176-256-5","work_id":"9545910e-7f35-4e07-849a-5e834bbb1f39","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Which side are you on? a multi-task dataset for end-to-end argument summarisation and evaluation","work_id":"9ac0939d-8c7c-4493-bfab-f2d99044ab67","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":12,"snapshot_sha256":"77e5d31555aebfa902f69974d4c032f4e0d9e48c0783cdb0d9d447713a0cb2e4","internal_anchors":2},"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":"2605.13793","created_at":"2026-05-18T02:44:15.596350+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.13793v1","created_at":"2026-05-18T02:44:15.596350+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13793","created_at":"2026-05-18T02:44:15.596350+00:00"},{"alias_kind":"pith_short_12","alias_value":"HVR6JYBJ7NDJ","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"HVR6JYBJ7NDJ5PFQ","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"HVR6JYBJ","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB","json":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB.json","graph_json":"https://pith.science/api/pith-number/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/graph.json","events_json":"https://pith.science/api/pith-number/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/events.json","paper":"https://pith.science/paper/HVR6JYBJ"},"agent_actions":{"view_html":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB","download_json":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB.json","view_paper":"https://pith.science/paper/HVR6JYBJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.13793&json=true","fetch_graph":"https://pith.science/api/pith-number/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/graph.json","fetch_events":"https://pith.science/api/pith-number/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/action/storage_attestation","attest_author":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/action/author_attestation","sign_citation":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/action/citation_signature","submit_replication":"https://pith.science/pith/HVR6JYBJ7NDJ5PFQG6V5U7EYMB/action/replication_record"}},"created_at":"2026-05-18T02:44:15.596350+00:00","updated_at":"2026-05-18T02:44:15.596350+00:00"}