{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:QDF5NZG7SPMJD6O7H2537YF7PU","short_pith_number":"pith:QDF5NZG7","schema_version":"1.0","canonical_sha256":"80cbd6e4df93d891f9df3ebbbfe0bf7d2ffabf0aa3e65d59c179c2e807faea11","source":{"kind":"arxiv","id":"2010.12834","version":2},"attestation_state":"computed","paper":{"title":"GO FIGURE: A Meta Evaluation of Factuality in Summarization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Asli Celikyilmaz, Jianfeng Gao, Rahul Jha, Saadia Gabriel, Yejin Choi","submitted_at":"2020-10-24T08:30:20Z","abstract_excerpt":"While neural language models can generate text with remarkable fluency and coherence, controlling for factual correctness in generation remains an open research question. This major discrepancy between the surface-level fluency and the content-level correctness of neural generation has motivated a new line of research that seeks automatic metrics for evaluating the factuality of machine text. In this paper, we introduce GO FIGURE, a meta-evaluation framework for evaluating factuality evaluation metrics. We propose five necessary and intuitive conditions to evaluate factuality metrics on diagno"},"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":"2010.12834","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-10-24T08:30:20Z","cross_cats_sorted":[],"title_canon_sha256":"4b1c3dae6e54e417427b8ae5bfb34cb0662ef08c5c001666d9a37defbc99e7c1","abstract_canon_sha256":"9435b2a2aeecfa115c50eb641b453cd323737b5d39febbbd4d7648578cea3517"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:46:25.152348Z","signature_b64":"vWTSmMZtUtSZ2PxSENhbtTngZCnXqsXPC2e7QrNmhUIBbKgtuV7bAlJfWEo71v0lfqztuFWpHtRSFAr7HB4uAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"80cbd6e4df93d891f9df3ebbbfe0bf7d2ffabf0aa3e65d59c179c2e807faea11","last_reissued_at":"2026-07-05T02:46:25.151841Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:46:25.151841Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GO FIGURE: A Meta Evaluation of Factuality in Summarization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Asli Celikyilmaz, Jianfeng Gao, Rahul Jha, Saadia Gabriel, Yejin Choi","submitted_at":"2020-10-24T08:30:20Z","abstract_excerpt":"While neural language models can generate text with remarkable fluency and coherence, controlling for factual correctness in generation remains an open research question. This major discrepancy between the surface-level fluency and the content-level correctness of neural generation has motivated a new line of research that seeks automatic metrics for evaluating the factuality of machine text. In this paper, we introduce GO FIGURE, a meta-evaluation framework for evaluating factuality evaluation metrics. We propose five necessary and intuitive conditions to evaluate factuality metrics on diagno"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.12834","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/2010.12834/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":"2010.12834","created_at":"2026-07-05T02:46:25.151899+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.12834v2","created_at":"2026-07-05T02:46:25.151899+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.12834","created_at":"2026-07-05T02:46:25.151899+00:00"},{"alias_kind":"pith_short_12","alias_value":"QDF5NZG7SPMJ","created_at":"2026-07-05T02:46:25.151899+00:00"},{"alias_kind":"pith_short_16","alias_value":"QDF5NZG7SPMJD6O7","created_at":"2026-07-05T02:46:25.151899+00:00"},{"alias_kind":"pith_short_8","alias_value":"QDF5NZG7","created_at":"2026-07-05T02:46:25.151899+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.20131","citing_title":"Whose Story Gets Told? Positionality and Bias in LLM Summaries of Life Narratives","ref_index":141,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU","json":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU.json","graph_json":"https://pith.science/api/pith-number/QDF5NZG7SPMJD6O7H2537YF7PU/graph.json","events_json":"https://pith.science/api/pith-number/QDF5NZG7SPMJD6O7H2537YF7PU/events.json","paper":"https://pith.science/paper/QDF5NZG7"},"agent_actions":{"view_html":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU","download_json":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU.json","view_paper":"https://pith.science/paper/QDF5NZG7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.12834&json=true","fetch_graph":"https://pith.science/api/pith-number/QDF5NZG7SPMJD6O7H2537YF7PU/graph.json","fetch_events":"https://pith.science/api/pith-number/QDF5NZG7SPMJD6O7H2537YF7PU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU/action/storage_attestation","attest_author":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU/action/author_attestation","sign_citation":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU/action/citation_signature","submit_replication":"https://pith.science/pith/QDF5NZG7SPMJD6O7H2537YF7PU/action/replication_record"}},"created_at":"2026-07-05T02:46:25.151899+00:00","updated_at":"2026-07-05T02:46:25.151899+00:00"}