{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:FHEZPXP5JC36QLGUWLNBEVJUMA","short_pith_number":"pith:FHEZPXP5","canonical_record":{"source":{"id":"2406.15227","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-21T15:11:33Z","cross_cats_sorted":[],"title_canon_sha256":"c2d33e05dc31660f27379cb94dc33c607ad7e87461ab9cee4f92bb05f4b3db8a","abstract_canon_sha256":"770e40d116cc0af27bcbe9bfaa23c611ae76a0fd51b49558df3567f91f2b1134"},"schema_version":"1.0"},"canonical_sha256":"29c997ddfd48b7e82cd4b2da125534600ed3ad3f8305af194393f9356ac2bd41","source":{"kind":"arxiv","id":"2406.15227","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.15227","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"arxiv_version","alias_value":"2406.15227v3","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.15227","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_12","alias_value":"FHEZPXP5JC36","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_16","alias_value":"FHEZPXP5JC36QLGU","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_8","alias_value":"FHEZPXP5","created_at":"2026-07-05T09:30:17Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:FHEZPXP5JC36QLGUWLNBEVJUMA","target":"record","payload":{"canonical_record":{"source":{"id":"2406.15227","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-21T15:11:33Z","cross_cats_sorted":[],"title_canon_sha256":"c2d33e05dc31660f27379cb94dc33c607ad7e87461ab9cee4f92bb05f4b3db8a","abstract_canon_sha256":"770e40d116cc0af27bcbe9bfaa23c611ae76a0fd51b49558df3567f91f2b1134"},"schema_version":"1.0"},"canonical_sha256":"29c997ddfd48b7e82cd4b2da125534600ed3ad3f8305af194393f9356ac2bd41","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:30:17.572636Z","signature_b64":"GOwcFNWKwGClua6KF5oCci9Vn9yrctzoWTHfSGEn3yMjYNAlxGQLr9Vv+lV4ABLVq54Ql7hs+9sH+wU1p8kZAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"29c997ddfd48b7e82cd4b2da125534600ed3ad3f8305af194393f9356ac2bd41","last_reissued_at":"2026-07-05T09:30:17.572141Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:30:17.572141Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.15227","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:30:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Em7Rc1Ov+oGSbZSQumLKtJoeGePcV1uut+tOMCzYHE2ZEAFRLbOQJhX17nVC6+4JjKVZiWo5j5BD40yHLef7Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T16:51:51.004714Z"},"content_sha256":"075ec13341b6f8a46e85e635fcc5cd5d542357fbf84c837689aed1f24e77e3fc","schema_version":"1.0","event_id":"sha256:075ec13341b6f8a46e85e635fcc5cd5d542357fbf84c837689aed1f24e77e3fc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:FHEZPXP5JC36QLGUWLNBEVJUMA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"A LLM-Based Ranking Method for the Evaluation of Automatic Counter-Narrative Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aitor Soroa, Irune Zubiaga, Rodrigo Agerri","submitted_at":"2024-06-21T15:11:33Z","abstract_excerpt":"This paper proposes a novel approach to evaluate Counter Narrative (CN) generation using a Large Language Model (LLM) as an evaluator. We show that traditional automatic metrics correlate poorly with human judgements and fail to capture the nuanced relationship between generated CNs and human perception. To alleviate this, we introduce a model ranking pipeline based on pairwise comparisons of generated CNs from different models, organized in a tournament-style format. The proposed evaluation method achieves a high correlation with human preference, with a $\\rho$ score of 0.88. As an additional"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.15227","kind":"arxiv","version":3},"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/2406.15227/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":"2026-07-05T09:30:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4o/OUrg4X3D4S8bwdm+dHni+s+/Scl8kbFFWpYqvsodkxqO4t4wNsBIVYgD10O2HCHSp4b9NU294SN/WvrghAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T16:51:51.005324Z"},"content_sha256":"837efae346ff674316a73153709eeb206e3051fee1ed208d26ddfe739c67bc46","schema_version":"1.0","event_id":"sha256:837efae346ff674316a73153709eeb206e3051fee1ed208d26ddfe739c67bc46"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FHEZPXP5JC36QLGUWLNBEVJUMA/bundle.json","state_url":"https://pith.science/pith/FHEZPXP5JC36QLGUWLNBEVJUMA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FHEZPXP5JC36QLGUWLNBEVJUMA/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":"2026-08-14T16:51:51Z","links":{"resolver":"https://pith.science/pith/FHEZPXP5JC36QLGUWLNBEVJUMA","bundle":"https://pith.science/pith/FHEZPXP5JC36QLGUWLNBEVJUMA/bundle.json","state":"https://pith.science/pith/FHEZPXP5JC36QLGUWLNBEVJUMA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FHEZPXP5JC36QLGUWLNBEVJUMA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:FHEZPXP5JC36QLGUWLNBEVJUMA","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":"770e40d116cc0af27bcbe9bfaa23c611ae76a0fd51b49558df3567f91f2b1134","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-21T15:11:33Z","title_canon_sha256":"c2d33e05dc31660f27379cb94dc33c607ad7e87461ab9cee4f92bb05f4b3db8a"},"schema_version":"1.0","source":{"id":"2406.15227","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.15227","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"arxiv_version","alias_value":"2406.15227v3","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.15227","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_12","alias_value":"FHEZPXP5JC36","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_16","alias_value":"FHEZPXP5JC36QLGU","created_at":"2026-07-05T09:30:17Z"},{"alias_kind":"pith_short_8","alias_value":"FHEZPXP5","created_at":"2026-07-05T09:30:17Z"}],"graph_snapshots":[{"event_id":"sha256:837efae346ff674316a73153709eeb206e3051fee1ed208d26ddfe739c67bc46","target":"graph","created_at":"2026-07-05T09:30:17Z","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/2406.15227/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper proposes a novel approach to evaluate Counter Narrative (CN) generation using a Large Language Model (LLM) as an evaluator. We show that traditional automatic metrics correlate poorly with human judgements and fail to capture the nuanced relationship between generated CNs and human perception. To alleviate this, we introduce a model ranking pipeline based on pairwise comparisons of generated CNs from different models, organized in a tournament-style format. The proposed evaluation method achieves a high correlation with human preference, with a $\\rho$ score of 0.88. As an additional","authors_text":"Aitor Soroa, Irune Zubiaga, Rodrigo Agerri","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-21T15:11:33Z","title":"A LLM-Based Ranking Method for the Evaluation of Automatic Counter-Narrative Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.15227","kind":"arxiv","version":3},"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":"sha256:075ec13341b6f8a46e85e635fcc5cd5d542357fbf84c837689aed1f24e77e3fc","target":"record","created_at":"2026-07-05T09:30:17Z","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":"770e40d116cc0af27bcbe9bfaa23c611ae76a0fd51b49558df3567f91f2b1134","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-21T15:11:33Z","title_canon_sha256":"c2d33e05dc31660f27379cb94dc33c607ad7e87461ab9cee4f92bb05f4b3db8a"},"schema_version":"1.0","source":{"id":"2406.15227","kind":"arxiv","version":3}},"canonical_sha256":"29c997ddfd48b7e82cd4b2da125534600ed3ad3f8305af194393f9356ac2bd41","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"29c997ddfd48b7e82cd4b2da125534600ed3ad3f8305af194393f9356ac2bd41","first_computed_at":"2026-07-05T09:30:17.572141Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:30:17.572141Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"GOwcFNWKwGClua6KF5oCci9Vn9yrctzoWTHfSGEn3yMjYNAlxGQLr9Vv+lV4ABLVq54Ql7hs+9sH+wU1p8kZAg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:30:17.572636Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.15227","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:075ec13341b6f8a46e85e635fcc5cd5d542357fbf84c837689aed1f24e77e3fc","sha256:837efae346ff674316a73153709eeb206e3051fee1ed208d26ddfe739c67bc46"],"state_sha256":"b0ecff1068f84e4c40ce558c801679f50ce4ae924d6f8440d3098826770c0380"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fB0tBfdKKY75WRa8yivEnWnb34qRU+fbcq1zbV8r60ydnAPy1SR2LXXoCXUOafpHzc5s63EWi1n7A3nFH/iODQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T16:51:51.012285Z","bundle_sha256":"23717ec772e984f0f3c1c06e2893b5e8cb81a3a7ca190d5dc677a614948c013f"}}