{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:QRFC3POY4SHLKT5N7ZCQO5K3BR","short_pith_number":"pith:QRFC3POY","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"},"canonical_sha256":"844a2dbdd8e48eb54fadfe4507755b0c5a5426a408e7a0716616d7d1bd4beb2c","source":{"kind":"arxiv","id":"2405.09935","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.09935","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"arxiv_version","alias_value":"2405.09935v2","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.09935","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"pith_short_12","alias_value":"QRFC3POY4SHL","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"pith_short_16","alias_value":"QRFC3POY4SHLKT5N","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"pith_short_8","alias_value":"QRFC3POY","created_at":"2026-07-05T08:22:40Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:QRFC3POY4SHLKT5N7ZCQO5K3BR","target":"record","payload":{"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"},"canonical_sha256":"844a2dbdd8e48eb54fadfe4507755b0c5a5426a408e7a0716616d7d1bd4beb2c","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"},"source_kind":"arxiv","source_id":"2405.09935","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:22:40Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7A1ywPknMyka2ug2JMPbCT87/fBcva4ZmkerM5xSpg17eDamFlcnuo15cx02kT9ovjz4G4goeO0nvPLsZEFcAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T09:17:55.686527Z"},"content_sha256":"66ddafd2dd4151fc699c08d9e0d132ad92276d26241128b69f7892757cef39c4","schema_version":"1.0","event_id":"sha256:66ddafd2dd4151fc699c08d9e0d132ad92276d26241128b69f7892757cef39c4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:QRFC3POY4SHLKT5N7ZCQO5K3BR","target":"graph","payload":{"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:22:40Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XW+llR6iVySMnBMNKg1+wChwyurD0xiDeY7FuppPnsNkPnn6pQYVB3EElBc8Fv2NU7vVkCfHqQVH5ASI2cyaDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T09:17:55.687105Z"},"content_sha256":"cbb4f8e054f7e1a0f9d843fb06361f16804afc876a8b52a1dd42095096115ff1","schema_version":"1.0","event_id":"sha256:cbb4f8e054f7e1a0f9d843fb06361f16804afc876a8b52a1dd42095096115ff1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/bundle.json","state_url":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-09T09:17:55Z","links":{"resolver":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR","bundle":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/bundle.json","state":"https://pith.science/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QRFC3POY4SHLKT5N7ZCQO5K3BR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:QRFC3POY4SHLKT5N7ZCQO5K3BR","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"9720408a6b4020b45185582f6c0328401fed119fc2ca645efe125007382ed921","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-16T09:41:12Z","title_canon_sha256":"00680078dd2d0d12baa5953f50b2fd59cab5c831a47b1e9544b6f1d21c941c53"},"schema_version":"1.0","source":{"id":"2405.09935","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.09935","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"arxiv_version","alias_value":"2405.09935v2","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.09935","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"pith_short_12","alias_value":"QRFC3POY4SHL","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"pith_short_16","alias_value":"QRFC3POY4SHLKT5N","created_at":"2026-07-05T08:22:40Z"},{"alias_kind":"pith_short_8","alias_value":"QRFC3POY","created_at":"2026-07-05T08:22:40Z"}],"graph_snapshots":[{"event_id":"sha256:cbb4f8e054f7e1a0f9d843fb06361f16804afc876a8b52a1dd42095096115ff1","target":"graph","created_at":"2026-07-05T08:22:40Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2405.09935/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"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","authors_text":"Alex Kim, Keonwoo Kim, Sangwon Yoon","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-16T09:41:12Z","title":"DEBATE: Devil's Advocate-Based Assessment and Text Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.09935","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:66ddafd2dd4151fc699c08d9e0d132ad92276d26241128b69f7892757cef39c4","target":"record","created_at":"2026-07-05T08:22:40Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"9720408a6b4020b45185582f6c0328401fed119fc2ca645efe125007382ed921","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-16T09:41:12Z","title_canon_sha256":"00680078dd2d0d12baa5953f50b2fd59cab5c831a47b1e9544b6f1d21c941c53"},"schema_version":"1.0","source":{"id":"2405.09935","kind":"arxiv","version":2}},"canonical_sha256":"844a2dbdd8e48eb54fadfe4507755b0c5a5426a408e7a0716616d7d1bd4beb2c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"844a2dbdd8e48eb54fadfe4507755b0c5a5426a408e7a0716616d7d1bd4beb2c","first_computed_at":"2026-07-05T08:22:40.614512Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:22:40.614512Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"UWfVyTP3Qg0L8nMqeifF1AAs8JyKR5DTnyQJhP2horx6qBMkvkBE5GHud53QUfySz3O4VCgIUxnUTKdyHp39BQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:22:40.614982Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.09935","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:66ddafd2dd4151fc699c08d9e0d132ad92276d26241128b69f7892757cef39c4","sha256:cbb4f8e054f7e1a0f9d843fb06361f16804afc876a8b52a1dd42095096115ff1"],"state_sha256":"223e60e7b286b6029093e821a8503df9bd474a4788163e55d24126e8d2440737"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HXKYuHShMRWz/6S5i7Yoc+Qke6ZZsTiq2srC1iegs92VawSgnMqDPdD2HdtJbQtn+cNWtw+RXC3yS0piezyiBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T09:17:55.691804Z","bundle_sha256":"d9f7e896e134f19badd9d79ca73f8c38f696ebcb89c2f8285d876fd6de21ab5a"}}