{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:34XX72FR6XQVNI5BFTB6HRM3CS","short_pith_number":"pith:34XX72FR","schema_version":"1.0","canonical_sha256":"df2f7fe8b1f5e156a3a12cc3e3c59b14a90a24898078d47f580e4c591105bea0","source":{"kind":"arxiv","id":"2312.12651","version":3},"attestation_state":"computed","paper":{"title":"Toxic Bias: Perspective API Misreads German as More Toxic","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SI","authors_text":"Francesco Pierri, Gianluca Nogara, Luca Luceri, Petter T\\\"ornberg, Silvia Giordano, Stefano Cresci","submitted_at":"2023-12-19T22:52:51Z","abstract_excerpt":"Proprietary public APIs play a crucial and growing role as research tools among social scientists. Among such APIs, Google's machine learning-based Perspective API is extensively utilized for assessing the toxicity of social media messages, providing both an important resource for researchers and automatic content moderation. However, this paper exposes an important bias in Perspective API concerning German language text. Through an in-depth examination of several datasets, we uncover intrinsic language biases within the multilingual model of Perspective API. We find that the toxicity assessme"},"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":"2312.12651","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SI","submitted_at":"2023-12-19T22:52:51Z","cross_cats_sorted":[],"title_canon_sha256":"164476a3e6f900c8e3f00fc8042a8dbfc0a38b050cca69935f2c3ebeb940e251","abstract_canon_sha256":"d3e1f048002d4bd5d0660dccaf97f4cac308d3aa6dff8a0a51bd50b81a453005"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:44:50.069867Z","signature_b64":"GHEibj3zhFHU2aQocu7GjvUnreQradwhi9lZ/5iXR3sWlA9ehmwIRpMFJX21cj9sM4q3Y85qBUS2IVGjFudgCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"df2f7fe8b1f5e156a3a12cc3e3c59b14a90a24898078d47f580e4c591105bea0","last_reissued_at":"2026-07-05T08:44:50.069384Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:44:50.069384Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Toxic Bias: Perspective API Misreads German as More Toxic","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.SI","authors_text":"Francesco Pierri, Gianluca Nogara, Luca Luceri, Petter T\\\"ornberg, Silvia Giordano, Stefano Cresci","submitted_at":"2023-12-19T22:52:51Z","abstract_excerpt":"Proprietary public APIs play a crucial and growing role as research tools among social scientists. Among such APIs, Google's machine learning-based Perspective API is extensively utilized for assessing the toxicity of social media messages, providing both an important resource for researchers and automatic content moderation. However, this paper exposes an important bias in Perspective API concerning German language text. Through an in-depth examination of several datasets, we uncover intrinsic language biases within the multilingual model of Perspective API. We find that the toxicity assessme"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.12651","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/2312.12651/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":"2312.12651","created_at":"2026-07-05T08:44:50.069443+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.12651v3","created_at":"2026-07-05T08:44:50.069443+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.12651","created_at":"2026-07-05T08:44:50.069443+00:00"},{"alias_kind":"pith_short_12","alias_value":"34XX72FR6XQV","created_at":"2026-07-05T08:44:50.069443+00:00"},{"alias_kind":"pith_short_16","alias_value":"34XX72FR6XQVNI5B","created_at":"2026-07-05T08:44:50.069443+00:00"},{"alias_kind":"pith_short_8","alias_value":"34XX72FR","created_at":"2026-07-05T08:44:50.069443+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.04975","citing_title":"Evaluating Chinese Large Language Models: The Influence of Persona Assignment on Stereotypes and Safeguards","ref_index":44,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS","json":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS.json","graph_json":"https://pith.science/api/pith-number/34XX72FR6XQVNI5BFTB6HRM3CS/graph.json","events_json":"https://pith.science/api/pith-number/34XX72FR6XQVNI5BFTB6HRM3CS/events.json","paper":"https://pith.science/paper/34XX72FR"},"agent_actions":{"view_html":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS","download_json":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS.json","view_paper":"https://pith.science/paper/34XX72FR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.12651&json=true","fetch_graph":"https://pith.science/api/pith-number/34XX72FR6XQVNI5BFTB6HRM3CS/graph.json","fetch_events":"https://pith.science/api/pith-number/34XX72FR6XQVNI5BFTB6HRM3CS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS/action/storage_attestation","attest_author":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS/action/author_attestation","sign_citation":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS/action/citation_signature","submit_replication":"https://pith.science/pith/34XX72FR6XQVNI5BFTB6HRM3CS/action/replication_record"}},"created_at":"2026-07-05T08:44:50.069443+00:00","updated_at":"2026-07-05T08:44:50.069443+00:00"}