{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:QRFC3POY4SHLKT5N7ZCQO5K3BR","short_pith_number":"pith:QRFC3POY","schema_version":"1.0","canonical_sha256":"844a2dbdd8e48eb54fadfe4507755b0c5a5426a408e7a0716616d7d1bd4beb2c","source":{"kind":"arxiv","id":"2405.09935","version":2},"attestation_state":"computed","paper":{"title":"DEBATE: Devil's Advocate-Based Assessment and Text Evaluation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Alex Kim, Keonwoo Kim, Sangwon Yoon","submitted_at":"2024-05-16T09:41:12Z","abstract_excerpt":"As natural language generation (NLG) models have become prevalent, systematically assessing the quality of machine-generated texts has become increasingly important. Recent studies introduce LLM-based evaluators that operate as reference-free metrics, demonstrating their capability to adeptly handle novel tasks. However, these models generally rely on a single-agent approach, which, we argue, introduces an inherent limit to their performance. This is because there exist biases in LLM agent's responses, including preferences for certain text structure or content. In this work, we propose DEBATE"},"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":"2405.09935","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-16T09:41:12Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"00680078dd2d0d12baa5953f50b2fd59cab5c831a47b1e9544b6f1d21c941c53","abstract_canon_sha256":"9720408a6b4020b45185582f6c0328401fed119fc2ca645efe125007382ed921"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:22:40.614982Z","signature_b64":"UWfVyTP3Qg0L8nMqeifF1AAs8JyKR5DTnyQJhP2horx6qBMkvkBE5GHud53QUfySz3O4VCgIUxnUTKdyHp39BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"844a2dbdd8e48eb54fadfe4507755b0c5a5426a408e7a0716616d7d1bd4beb2c","last_reissued_at":"2026-07-05T08:22:40.614512Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:22:40.614512Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DEBATE: Devil's Advocate-Based Assessment and Text Evaluation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Alex Kim, Keonwoo Kim, Sangwon Yoon","submitted_at":"2024-05-16T09:41:12Z","abstract_excerpt":"As natural language generation (NLG) models have become prevalent, systematically assessing the quality of machine-generated texts has become increasingly important. Recent studies introduce LLM-based evaluators that operate as reference-free metrics, demonstrating their capability to adeptly handle novel tasks. However, these models generally rely on a single-agent approach, which, we argue, introduces an inherent limit to their performance. This is because there exist biases in LLM agent's responses, including preferences for certain text structure or content. In this work, we propose DEBATE"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.09935","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/2405.09935/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":"2405.09935","created_at":"2026-07-05T08:22:40.614564+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.09935v2","created_at":"2026-07-05T08:22:40.614564+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.09935","created_at":"2026-07-05T08:22:40.614564+00:00"},{"alias_kind":"pith_short_12","alias_value":"QRFC3POY4SHL","created_at":"2026-07-05T08:22:40.614564+00:00"},{"alias_kind":"pith_short_16","alias_value":"QRFC3POY4SHLKT5N","created_at":"2026-07-05T08:22:40.614564+00:00"},{"alias_kind":"pith_short_8","alias_value":"QRFC3POY","created_at":"2026-07-05T08:22:40.614564+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.00178","citing_title":"Tournament of Prompts: Evolving LLM Instructions Through Structured Debates and Elo Ratings","ref_index":31,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR","json":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR.json","graph_json":"https://pith.science/api/pith-number/QRFC3POY4SHLKT5N7ZCQO5K3BR/graph.json","events_json":"https://pith.science/api/pith-number/QRFC3POY4SHLKT5N7ZCQO5K3BR/events.json","paper":"https://pith.science/paper/QRFC3POY"},"agent_actions":{"view_html":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR","download_json":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR.json","view_paper":"https://pith.science/paper/QRFC3POY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.09935&json=true","fetch_graph":"https://pith.science/api/pith-number/QRFC3POY4SHLKT5N7ZCQO5K3BR/graph.json","fetch_events":"https://pith.science/api/pith-number/QRFC3POY4SHLKT5N7ZCQO5K3BR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/action/storage_attestation","attest_author":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/action/author_attestation","sign_citation":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/action/citation_signature","submit_replication":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/action/replication_record"}},"created_at":"2026-07-05T08:22:40.614564+00:00","updated_at":"2026-07-05T08:22:40.614564+00:00"}