{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3TXYMSSYNKNTPHNALVLPMEIVPK","short_pith_number":"pith:3TXYMSSY","schema_version":"1.0","canonical_sha256":"dcef864a586a9b379da05d56f611157aaa3538578415cab73cb671a616f4507b","source":{"kind":"arxiv","id":"2606.00334","version":1},"attestation_state":"computed","paper":{"title":"Isolating LLM Lexical Bias: A Curation-Free Triangulated Metric for Preference-Stage Learning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jose Hernandez, Thomas Stephan Juzek, Xiaoyang Ming","submitted_at":"2026-05-29T20:19:49Z","abstract_excerpt":"Various language domains have undergone remarkable changes in recent years; these shifts are largely attributed to the advent of Large Language Models and their misalignment with natural language usage. These misalignments are thought to partly originate in the preference-learning stage, e.g. Reinforcement Learning from Human Feedback, which generally makes models more useful but simultaneously may introduce systematic lexical bias. In terms of lexical behavior, this is visible in a model's preference for certain formats or the overuse of words (delve, furthermore), even when such patterns are"},"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":"2606.00334","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-29T20:19:49Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"898b5630fac85ede725ea851438acbcc87117538e126decc8620bc1a75a63029","abstract_canon_sha256":"a94ae8da78e8b135a239cf4ba178ca4b408fc5e7b051f09d2023b8596a23b7f1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:51.904665Z","signature_b64":"CFxlUXFwSEfe7OHIjItkwuw6H9cuevzQfzyuJxD+bYHynVJHtwSq9jTNe+tMZKDDOWL97IGIy0X1vQId7oXEAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dcef864a586a9b379da05d56f611157aaa3538578415cab73cb671a616f4507b","last_reissued_at":"2026-06-02T01:03:51.904259Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:51.904259Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Isolating LLM Lexical Bias: A Curation-Free Triangulated Metric for Preference-Stage Learning","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jose Hernandez, Thomas Stephan Juzek, Xiaoyang Ming","submitted_at":"2026-05-29T20:19:49Z","abstract_excerpt":"Various language domains have undergone remarkable changes in recent years; these shifts are largely attributed to the advent of Large Language Models and their misalignment with natural language usage. These misalignments are thought to partly originate in the preference-learning stage, e.g. Reinforcement Learning from Human Feedback, which generally makes models more useful but simultaneously may introduce systematic lexical bias. In terms of lexical behavior, this is visible in a model's preference for certain formats or the overuse of words (delve, furthermore), even when such patterns are"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.00334","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/2606.00334/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":"2606.00334","created_at":"2026-06-02T01:03:51.904323+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.00334v1","created_at":"2026-06-02T01:03:51.904323+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.00334","created_at":"2026-06-02T01:03:51.904323+00:00"},{"alias_kind":"pith_short_12","alias_value":"3TXYMSSYNKNT","created_at":"2026-06-02T01:03:51.904323+00:00"},{"alias_kind":"pith_short_16","alias_value":"3TXYMSSYNKNTPHNA","created_at":"2026-06-02T01:03:51.904323+00:00"},{"alias_kind":"pith_short_8","alias_value":"3TXYMSSY","created_at":"2026-06-02T01:03:51.904323+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/3TXYMSSYNKNTPHNALVLPMEIVPK","json":"https://pith.science/pith/3TXYMSSYNKNTPHNALVLPMEIVPK.json","graph_json":"https://pith.science/api/pith-number/3TXYMSSYNKNTPHNALVLPMEIVPK/graph.json","events_json":"https://pith.science/api/pith-number/3TXYMSSYNKNTPHNALVLPMEIVPK/events.json","paper":"https://pith.science/paper/3TXYMSSY"},"agent_actions":{"view_html":"https://pith.science/pith/3TXYMSSYNKNTPHNALVLPMEIVPK","download_json":"https://pith.science/pith/3TXYMSSYNKNTPHNALVLPMEIVPK.json","view_paper":"https://pith.science/paper/3TXYMSSY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.00334&json=true","fetch_graph":"https://pith.science/api/pith-number/3TXYMSSYNKNTPHNALVLPMEIVPK/graph.json","fetch_events":"https://pith.science/api/pith-number/3TXYMSSYNKNTPHNALVLPMEIVPK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3TXYMSSYNKNTPHNALVLPMEIVPK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3TXYMSSYNKNTPHNALVLPMEIVPK/action/storage_attestation","attest_author":"https://pith.science/pith/3TXYMSSYNKNTPHNALVLPMEIVPK/action/author_attestation","sign_citation":"https://pith.science/pith/3TXYMSSYNKNTPHNALVLPMEIVPK/action/citation_signature","submit_replication":"https://pith.science/pith/3TXYMSSYNKNTPHNALVLPMEIVPK/action/replication_record"}},"created_at":"2026-06-02T01:03:51.904323+00:00","updated_at":"2026-06-02T01:03:51.904323+00:00"}