{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:ZCJLKREYXQAZILWFWBUZPRLIMW","short_pith_number":"pith:ZCJLKREY","schema_version":"1.0","canonical_sha256":"c892b54498bc01942ec5b06997c56865ac7a6055db368edb7200e2ae2cd66e09","source":{"kind":"arxiv","id":"2310.16959","version":1},"attestation_state":"computed","paper":{"title":"Improving Few-shot Generalization of Safety Classifiers via Data Augmented Parameter-Efficient Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ahmad Beirami, Alex Beutel, Ananth Balashankar, Aradhana Sinha, Jilin Chen, Xiao Ma, Yao Qin","submitted_at":"2023-10-25T19:57:07Z","abstract_excerpt":"As large language models (LLMs) are widely adopted, new safety issues and policies emerge, to which existing safety classifiers do not generalize well. If we have only observed a few examples of violations of a new safety rule, how can we build a classifier to detect violations? In this paper, we study the novel setting of domain-generalized few-shot learning for LLM-based text safety classifiers. Unlike prior few-shot work, these new safety issues can be hard to uncover and we do not get to choose the few examples. We demonstrate that existing few-shot techniques do not perform well in this s"},"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":"2310.16959","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-25T19:57:07Z","cross_cats_sorted":[],"title_canon_sha256":"d4680daef9b6b6485c8c50dec13f412b3b0fc0789d4b035d69114be0f5a1a6eb","abstract_canon_sha256":"e517cdbea07fe010384fe6c188c26aa0b2f33b2c02ddeadde3d9bd04201c7a28"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:05:16.607763Z","signature_b64":"gKw+gn4q73985RTxyeVTNqT5c/UCIyFXDBRia3a54wEfKi3W7h0s+LjeZnGAH7Lc6Ajy7uiNfZBKnICzAqw2Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c892b54498bc01942ec5b06997c56865ac7a6055db368edb7200e2ae2cd66e09","last_reissued_at":"2026-07-05T07:05:16.607186Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:05:16.607186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improving Few-shot Generalization of Safety Classifiers via Data Augmented Parameter-Efficient Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ahmad Beirami, Alex Beutel, Ananth Balashankar, Aradhana Sinha, Jilin Chen, Xiao Ma, Yao Qin","submitted_at":"2023-10-25T19:57:07Z","abstract_excerpt":"As large language models (LLMs) are widely adopted, new safety issues and policies emerge, to which existing safety classifiers do not generalize well. If we have only observed a few examples of violations of a new safety rule, how can we build a classifier to detect violations? In this paper, we study the novel setting of domain-generalized few-shot learning for LLM-based text safety classifiers. Unlike prior few-shot work, these new safety issues can be hard to uncover and we do not get to choose the few examples. We demonstrate that existing few-shot techniques do not perform well in this s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.16959","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/2310.16959/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":"2310.16959","created_at":"2026-07-05T07:05:16.607255+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.16959v1","created_at":"2026-07-05T07:05:16.607255+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.16959","created_at":"2026-07-05T07:05:16.607255+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZCJLKREYXQAZ","created_at":"2026-07-05T07:05:16.607255+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZCJLKREYXQAZILWF","created_at":"2026-07-05T07:05:16.607255+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZCJLKREY","created_at":"2026-07-05T07:05:16.607255+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/ZCJLKREYXQAZILWFWBUZPRLIMW","json":"https://pith.science/pith/ZCJLKREYXQAZILWFWBUZPRLIMW.json","graph_json":"https://pith.science/api/pith-number/ZCJLKREYXQAZILWFWBUZPRLIMW/graph.json","events_json":"https://pith.science/api/pith-number/ZCJLKREYXQAZILWFWBUZPRLIMW/events.json","paper":"https://pith.science/paper/ZCJLKREY"},"agent_actions":{"view_html":"https://pith.science/pith/ZCJLKREYXQAZILWFWBUZPRLIMW","download_json":"https://pith.science/pith/ZCJLKREYXQAZILWFWBUZPRLIMW.json","view_paper":"https://pith.science/paper/ZCJLKREY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.16959&json=true","fetch_graph":"https://pith.science/api/pith-number/ZCJLKREYXQAZILWFWBUZPRLIMW/graph.json","fetch_events":"https://pith.science/api/pith-number/ZCJLKREYXQAZILWFWBUZPRLIMW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZCJLKREYXQAZILWFWBUZPRLIMW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZCJLKREYXQAZILWFWBUZPRLIMW/action/storage_attestation","attest_author":"https://pith.science/pith/ZCJLKREYXQAZILWFWBUZPRLIMW/action/author_attestation","sign_citation":"https://pith.science/pith/ZCJLKREYXQAZILWFWBUZPRLIMW/action/citation_signature","submit_replication":"https://pith.science/pith/ZCJLKREYXQAZILWFWBUZPRLIMW/action/replication_record"}},"created_at":"2026-07-05T07:05:16.607255+00:00","updated_at":"2026-07-05T07:05:16.607255+00:00"}