{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:E4GRGVJVIZ5ZK5AAO4IN4V4HT7","short_pith_number":"pith:E4GRGVJV","schema_version":"1.0","canonical_sha256":"270d135535467b9574007710de57879fea3e9a85eec0e01f94f1b89f99683618","source":{"kind":"arxiv","id":"2203.11131","version":1},"attestation_state":"computed","paper":{"title":"Towards Explainable Evaluation Metrics for Natural Language Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Christoph Leiter, Marina Fomicheva, Piyawat Lertvittayakumjorn, Steffen Eger, Wei Zhao, Yang Gao","submitted_at":"2022-03-21T17:05:54Z","abstract_excerpt":"Unlike classical lexical overlap metrics such as BLEU, most current evaluation metrics (such as BERTScore or MoverScore) are based on black-box language models such as BERT or XLM-R. They often achieve strong correlations with human judgments, but recent research indicates that the lower-quality classical metrics remain dominant, one of the potential reasons being that their decision processes are transparent. To foster more widespread acceptance of the novel high-quality metrics, explainability thus becomes crucial. In this concept paper, we identify key properties and propose key goals of ex"},"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":"2203.11131","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-03-21T17:05:54Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"58a91a92613b3bd322eb44341289415b1b53840d3778ab12dd165e2472f851da","abstract_canon_sha256":"09775161148d80a3eba4677266a267d6c759800504cbe68cb250b5feb50c0907"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:07:00.969824Z","signature_b64":"0oTuYM0KPaOoQAj6cbRGGWXTvir8zTht5lZKaYykYamHBvHYuFAVCHQ29bAIJvRWeYjIDgWlL77UJTBT4kE3BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"270d135535467b9574007710de57879fea3e9a85eec0e01f94f1b89f99683618","last_reissued_at":"2026-07-05T04:07:00.969368Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:07:00.969368Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Explainable Evaluation Metrics for Natural Language Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Christoph Leiter, Marina Fomicheva, Piyawat Lertvittayakumjorn, Steffen Eger, Wei Zhao, Yang Gao","submitted_at":"2022-03-21T17:05:54Z","abstract_excerpt":"Unlike classical lexical overlap metrics such as BLEU, most current evaluation metrics (such as BERTScore or MoverScore) are based on black-box language models such as BERT or XLM-R. They often achieve strong correlations with human judgments, but recent research indicates that the lower-quality classical metrics remain dominant, one of the potential reasons being that their decision processes are transparent. To foster more widespread acceptance of the novel high-quality metrics, explainability thus becomes crucial. In this concept paper, we identify key properties and propose key goals of ex"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.11131","kind":"arxiv","version":1},"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/2203.11131/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":"2203.11131","created_at":"2026-07-05T04:07:00.969428+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.11131v1","created_at":"2026-07-05T04:07:00.969428+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.11131","created_at":"2026-07-05T04:07:00.969428+00:00"},{"alias_kind":"pith_short_12","alias_value":"E4GRGVJVIZ5Z","created_at":"2026-07-05T04:07:00.969428+00:00"},{"alias_kind":"pith_short_16","alias_value":"E4GRGVJVIZ5ZK5AA","created_at":"2026-07-05T04:07:00.969428+00:00"},{"alias_kind":"pith_short_8","alias_value":"E4GRGVJV","created_at":"2026-07-05T04:07:00.969428+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.14956","citing_title":"ExPerT: Effective and Explainable Evaluation of Personalized Long-Form Text Generation","ref_index":23,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7","json":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7.json","graph_json":"https://pith.science/api/pith-number/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/graph.json","events_json":"https://pith.science/api/pith-number/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/events.json","paper":"https://pith.science/paper/E4GRGVJV"},"agent_actions":{"view_html":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7","download_json":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7.json","view_paper":"https://pith.science/paper/E4GRGVJV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.11131&json=true","fetch_graph":"https://pith.science/api/pith-number/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/graph.json","fetch_events":"https://pith.science/api/pith-number/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/action/storage_attestation","attest_author":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/action/author_attestation","sign_citation":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/action/citation_signature","submit_replication":"https://pith.science/pith/E4GRGVJVIZ5ZK5AAO4IN4V4HT7/action/replication_record"}},"created_at":"2026-07-05T04:07:00.969428+00:00","updated_at":"2026-07-05T04:07:00.969428+00:00"}