{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:TUJBK2LISCMPH7I4M62SPR2VGR","short_pith_number":"pith:TUJBK2LI","schema_version":"1.0","canonical_sha256":"9d121569689098f3fd1c67b527c7553462decdd8594a7e58a1f84c9355ede630","source":{"kind":"arxiv","id":"2411.10954","version":1},"attestation_state":"computed","paper":{"title":"Dialectal Toxicity Detection: Evaluating LLM-as-a-Judge Consistency Across Language Varieties","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Antonios Anastasopoulos, Fahim Faisal, Md Mushfiqur Rahman","submitted_at":"2024-11-17T03:53:24Z","abstract_excerpt":"There has been little systematic study on how dialectal differences affect toxicity detection by modern LLMs. Furthermore, although using LLMs as evaluators (\"LLM-as-a-judge\") is a growing research area, their sensitivity to dialectal nuances is still underexplored and requires more focused attention. In this paper, we address these gaps through a comprehensive toxicity evaluation of LLMs across diverse dialects. We create a multi-dialect dataset through synthetic transformations and human-assisted translations, covering 10 language clusters and 60 varieties. We then evaluated three LLMs on th"},"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":"2411.10954","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-17T03:53:24Z","cross_cats_sorted":[],"title_canon_sha256":"be08ef03448323d07204ed0a9824871e9d89dac4e811e8b6818155f855ea2b2a","abstract_canon_sha256":"f1b27f1a25635fd7ab99b8fc65eb91b97fd66805211aa8ef42a423022cb5d8ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:36:26.004419Z","signature_b64":"CP6o6Ip6uZv/C6HmFe2MByi3hvBtLbaLPgpJ9Wcj4GuRwsu8iLhIHsjmijJT4rTgJcFQkiOh2XAeRxt3RglkDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9d121569689098f3fd1c67b527c7553462decdd8594a7e58a1f84c9355ede630","last_reissued_at":"2026-07-05T09:36:26.003929Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:36:26.003929Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dialectal Toxicity Detection: Evaluating LLM-as-a-Judge Consistency Across Language Varieties","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Antonios Anastasopoulos, Fahim Faisal, Md Mushfiqur Rahman","submitted_at":"2024-11-17T03:53:24Z","abstract_excerpt":"There has been little systematic study on how dialectal differences affect toxicity detection by modern LLMs. Furthermore, although using LLMs as evaluators (\"LLM-as-a-judge\") is a growing research area, their sensitivity to dialectal nuances is still underexplored and requires more focused attention. In this paper, we address these gaps through a comprehensive toxicity evaluation of LLMs across diverse dialects. We create a multi-dialect dataset through synthetic transformations and human-assisted translations, covering 10 language clusters and 60 varieties. We then evaluated three LLMs on th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.10954","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/2411.10954/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":"2411.10954","created_at":"2026-07-05T09:36:26.003991+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.10954v1","created_at":"2026-07-05T09:36:26.003991+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.10954","created_at":"2026-07-05T09:36:26.003991+00:00"},{"alias_kind":"pith_short_12","alias_value":"TUJBK2LISCMP","created_at":"2026-07-05T09:36:26.003991+00:00"},{"alias_kind":"pith_short_16","alias_value":"TUJBK2LISCMPH7I4","created_at":"2026-07-05T09:36:26.003991+00:00"},{"alias_kind":"pith_short_8","alias_value":"TUJBK2LI","created_at":"2026-07-05T09:36:26.003991+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.00794","citing_title":"Measurement Without Validity: The Compounding Reliability Problem in Agentic AI Evaluation","ref_index":32,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR","json":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR.json","graph_json":"https://pith.science/api/pith-number/TUJBK2LISCMPH7I4M62SPR2VGR/graph.json","events_json":"https://pith.science/api/pith-number/TUJBK2LISCMPH7I4M62SPR2VGR/events.json","paper":"https://pith.science/paper/TUJBK2LI"},"agent_actions":{"view_html":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR","download_json":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR.json","view_paper":"https://pith.science/paper/TUJBK2LI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.10954&json=true","fetch_graph":"https://pith.science/api/pith-number/TUJBK2LISCMPH7I4M62SPR2VGR/graph.json","fetch_events":"https://pith.science/api/pith-number/TUJBK2LISCMPH7I4M62SPR2VGR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR/action/storage_attestation","attest_author":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR/action/author_attestation","sign_citation":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR/action/citation_signature","submit_replication":"https://pith.science/pith/TUJBK2LISCMPH7I4M62SPR2VGR/action/replication_record"}},"created_at":"2026-07-05T09:36:26.003991+00:00","updated_at":"2026-07-05T09:36:26.003991+00:00"}