{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:27P3IBH5PVHACZUSYJYSYEWBSA","short_pith_number":"pith:27P3IBH5","schema_version":"1.0","canonical_sha256":"d7dfb404fd7d4e016692c2712c12c19030adb75031fe7d65a0ca10604acd428b","source":{"kind":"arxiv","id":"2210.07626","version":1},"attestation_state":"computed","paper":{"title":"BERTScore is Unfair: On Social Bias in Language Model-Based Metrics for Text Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Junliang He, Tianxiang Sun, Xipeng Qiu, Xuanjing Huang","submitted_at":"2022-10-14T08:24:11Z","abstract_excerpt":"Automatic evaluation metrics are crucial to the development of generative systems. In recent years, pre-trained language model (PLM) based metrics, such as BERTScore, have been commonly adopted in various generation tasks. However, it has been demonstrated that PLMs encode a range of stereotypical societal biases, leading to a concern on the fairness of PLMs as metrics. To that end, this work presents the first systematic study on the social bias in PLM-based metrics. We demonstrate that popular PLM-based metrics exhibit significantly higher social bias than traditional metrics on 6 sensitive "},"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":"2210.07626","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-10-14T08:24:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a3f0b10e2b867e209aa7ffafb8e88312d7fab24c84f8eb4f916ad795e8cda529","abstract_canon_sha256":"0f76956486d485ae9d6f92aefde0314250e63ce34e1528cb7f8cfbea7880d9ea"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:06:39.439387Z","signature_b64":"Bb71txcwefSE3e6vMgJy2UzsdbCtifr8AzQGbMoLYzLiBxHCHcNgSJPOpt/epnL0/phtZHG5o8yZOBBca8w1CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d7dfb404fd7d4e016692c2712c12c19030adb75031fe7d65a0ca10604acd428b","last_reissued_at":"2026-07-05T05:06:39.438998Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:06:39.438998Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BERTScore is Unfair: On Social Bias in Language Model-Based Metrics for Text Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Junliang He, Tianxiang Sun, Xipeng Qiu, Xuanjing Huang","submitted_at":"2022-10-14T08:24:11Z","abstract_excerpt":"Automatic evaluation metrics are crucial to the development of generative systems. In recent years, pre-trained language model (PLM) based metrics, such as BERTScore, have been commonly adopted in various generation tasks. However, it has been demonstrated that PLMs encode a range of stereotypical societal biases, leading to a concern on the fairness of PLMs as metrics. To that end, this work presents the first systematic study on the social bias in PLM-based metrics. We demonstrate that popular PLM-based metrics exhibit significantly higher social bias than traditional metrics on 6 sensitive "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.07626","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/2210.07626/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":"2210.07626","created_at":"2026-07-05T05:06:39.439053+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.07626v1","created_at":"2026-07-05T05:06:39.439053+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.07626","created_at":"2026-07-05T05:06:39.439053+00:00"},{"alias_kind":"pith_short_12","alias_value":"27P3IBH5PVHA","created_at":"2026-07-05T05:06:39.439053+00:00"},{"alias_kind":"pith_short_16","alias_value":"27P3IBH5PVHACZUS","created_at":"2026-07-05T05:06:39.439053+00:00"},{"alias_kind":"pith_short_8","alias_value":"27P3IBH5","created_at":"2026-07-05T05:06:39.439053+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA","json":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA.json","graph_json":"https://pith.science/api/pith-number/27P3IBH5PVHACZUSYJYSYEWBSA/graph.json","events_json":"https://pith.science/api/pith-number/27P3IBH5PVHACZUSYJYSYEWBSA/events.json","paper":"https://pith.science/paper/27P3IBH5"},"agent_actions":{"view_html":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA","download_json":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA.json","view_paper":"https://pith.science/paper/27P3IBH5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.07626&json=true","fetch_graph":"https://pith.science/api/pith-number/27P3IBH5PVHACZUSYJYSYEWBSA/graph.json","fetch_events":"https://pith.science/api/pith-number/27P3IBH5PVHACZUSYJYSYEWBSA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA/action/storage_attestation","attest_author":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA/action/author_attestation","sign_citation":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA/action/citation_signature","submit_replication":"https://pith.science/pith/27P3IBH5PVHACZUSYJYSYEWBSA/action/replication_record"}},"created_at":"2026-07-05T05:06:39.439053+00:00","updated_at":"2026-07-05T05:06:39.439053+00:00"}