{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:3A3NSDLDDAURSZEKB2WUDWSEKN","short_pith_number":"pith:3A3NSDLD","canonical_record":{"source":{"id":"2605.13538","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-13T13:47:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"71ca1095213f9418fa258a120300c4beee4c761bb6f28c94731d6b266582527e","abstract_canon_sha256":"08982f296d3b7011c8b7b2fdf6bec4c5c580d305c1cc7386049148fae213f8dc"},"schema_version":"1.0"},"canonical_sha256":"d836d90d63182919648a0ead41da445340ce937cfd128a535c945fe435ff69ef","source":{"kind":"arxiv","id":"2605.13538","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13538","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13538v1","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13538","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"pith_short_12","alias_value":"3A3NSDLDDAUR","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"3A3NSDLDDAURSZEK","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"3A3NSDLD","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:3A3NSDLDDAURSZEKB2WUDWSEKN","target":"record","payload":{"canonical_record":{"source":{"id":"2605.13538","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-13T13:47:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"71ca1095213f9418fa258a120300c4beee4c761bb6f28c94731d6b266582527e","abstract_canon_sha256":"08982f296d3b7011c8b7b2fdf6bec4c5c580d305c1cc7386049148fae213f8dc"},"schema_version":"1.0"},"canonical_sha256":"d836d90d63182919648a0ead41da445340ce937cfd128a535c945fe435ff69ef","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:24.089214Z","signature_b64":"UiiGXX66GqPyk1m3keJIB02P+V1lkUMgi/TVIISJ75wsBSzxgqypbEdhHkYG6O7UB9AclxzmKuuQDANrTUErDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d836d90d63182919648a0ead41da445340ce937cfd128a535c945fe435ff69ef","last_reissued_at":"2026-05-18T02:44:24.088768Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:24.088768Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.13538","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ANH+319vxfodIeJGLjrsYiUqf3AuDl5taIzVPUADMsNQx8SME/maBF3dFOx4VlR9wp5QZbIqML98t77RMTi8CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-19T16:31:48.438889Z"},"content_sha256":"e205aef49527f7367491c5b765d6e765333d1476e8829e1eb8aab59b63cfaba6","schema_version":"1.0","event_id":"sha256:e205aef49527f7367491c5b765d6e765333d1476e8829e1eb8aab59b63cfaba6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:3A3NSDLDDAURSZEKB2WUDWSEKN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Locale-Conditioned Few-Shot Prompting Mitigates Demonstration Regurgitation in On-Device PII Substitution with Small Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Locale-conditioned rotating few-shot prompts stop small language models from echoing demonstration examples during on-device PII substitution.","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Anuj Sadani, Deepak Kumar","submitted_at":"2026-05-13T13:47:11Z","abstract_excerpt":"Personally Identifiable Information (PII) redaction usually replaces detected entities with placeholder tokens such as [PERSON], destroying the downstream utility of the redacted text for retrieval and Named Entity Recognition (NER) training. We propose a fully on-device pipeline that substitutes PII with consistent, type-preserving fake values: a 1.5 B mixture-of-experts token classifier (openai/privacy-filter) detects spans, a 1-bit Bonsai-1.7B Small Language Model (SLM) proposes contextual surrogates for names, addresses, and dates, and a rule-based generator (faker) handles patterned field"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"With the fix, 482/482 unique Bonsai-1.7B calls succeed (no echoes) and produce locale-correct surrogates, although the SLM still copies from a small same-locale demonstration pool - a residual narrowness we quantify.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the observed NER performance gap is caused primarily by reduced variety in SLM outputs rather than by other unmeasured differences in the 160/40 subset or by the choice of XGLM-564M as the multilingual evaluator.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Locale-conditioned rotating few-shot prompting eliminates demonstration regurgitation in 1.7B SLMs for PII substitution while producing more natural text than rule-based methods, though downstream NER training benefits more from synthetic variety than naturalness.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Locale-conditioned rotating few-shot prompts stop small language models from echoing demonstration examples during on-device PII substitution.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"1767d0d1798e97fd46bb4e9176590148e9eb44ae983ad5d85fb458dc47ca6388"},"source":{"id":"2605.13538","kind":"arxiv","version":1},"verdict":{"id":"fddab2d7-c488-4d6a-8201-d6bc53eabbb7","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:00:57.673618Z","strongest_claim":"With the fix, 482/482 unique Bonsai-1.7B calls succeed (no echoes) and produce locale-correct surrogates, although the SLM still copies from a small same-locale demonstration pool - a residual narrowness we quantify.","one_line_summary":"Locale-conditioned rotating few-shot prompting eliminates demonstration regurgitation in 1.7B SLMs for PII substitution while producing more natural text than rule-based methods, though downstream NER training benefits more from synthetic variety than naturalness.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the observed NER performance gap is caused primarily by reduced variety in SLM outputs rather than by other unmeasured differences in the 160/40 subset or by the choice of XGLM-564M as the multilingual evaluator.","pith_extraction_headline":"Locale-conditioned rotating few-shot prompts stop small language models from echoing demonstration examples during on-device PII substitution."},"references":{"count":16,"sample":[{"doi":"","year":2024,"title":"Faker: a Python package that generates fake data","work_id":"b1918f04-be18-4cdc-8b1b-8098eac3c9ec","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"llama.cpp: Port of LLaMA models in C/C++","work_id":"6b334f6b-fac2-4e0c-910e-43ed4197b5c2","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"spaCy: Industrial-strength natural language processing in Python","work_id":"8878703d-2330-47d2-81ec-6f3b4c7eacb6","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Few-shot learning with multilingual language models","work_id":"32d56185-d394-41f6-a038-08c943a4650c","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Fantastically ordered prompts and where to find them: Overcoming few-shot prompt order sensitivity","work_id":"c6ecb82b-59ca-437f-91d8-470531f9f7c8","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":16,"snapshot_sha256":"8f7fc02398e1eadab2ea9d9fadc8e255a1580afbb705951483156df5e9d4b78b","internal_anchors":3},"formal_canon":{"evidence_count":1,"snapshot_sha256":"13cea1224d6f07fff5395eb607961feb0d0eb9b58ef3be5215275e4827c88642"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"fddab2d7-c488-4d6a-8201-d6bc53eabbb7"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"39N6rn0e4Q+qS9ijR3JYro2TPsODQ86keOFOqADbBeXCjjW94ktb8uJf2M5MdlACsqMCbnUBXQ6iDDNSpqNDAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-19T16:31:48.439453Z"},"content_sha256":"9aeb073cd6a4ee78316646a7a496d8c3908964698b74e9f10aa92be23816e870","schema_version":"1.0","event_id":"sha256:9aeb073cd6a4ee78316646a7a496d8c3908964698b74e9f10aa92be23816e870"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3A3NSDLDDAURSZEKB2WUDWSEKN/bundle.json","state_url":"https://pith.science/pith/3A3NSDLDDAURSZEKB2WUDWSEKN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3A3NSDLDDAURSZEKB2WUDWSEKN/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-19T16:31:48Z","links":{"resolver":"https://pith.science/pith/3A3NSDLDDAURSZEKB2WUDWSEKN","bundle":"https://pith.science/pith/3A3NSDLDDAURSZEKB2WUDWSEKN/bundle.json","state":"https://pith.science/pith/3A3NSDLDDAURSZEKB2WUDWSEKN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3A3NSDLDDAURSZEKB2WUDWSEKN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:3A3NSDLDDAURSZEKB2WUDWSEKN","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"08982f296d3b7011c8b7b2fdf6bec4c5c580d305c1cc7386049148fae213f8dc","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-13T13:47:11Z","title_canon_sha256":"71ca1095213f9418fa258a120300c4beee4c761bb6f28c94731d6b266582527e"},"schema_version":"1.0","source":{"id":"2605.13538","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13538","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13538v1","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13538","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"pith_short_12","alias_value":"3A3NSDLDDAUR","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"3A3NSDLDDAURSZEK","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"3A3NSDLD","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:9aeb073cd6a4ee78316646a7a496d8c3908964698b74e9f10aa92be23816e870","target":"graph","created_at":"2026-05-18T02:44:24Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"With the fix, 482/482 unique Bonsai-1.7B calls succeed (no echoes) and produce locale-correct surrogates, although the SLM still copies from a small same-locale demonstration pool - a residual narrowness we quantify."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the observed NER performance gap is caused primarily by reduced variety in SLM outputs rather than by other unmeasured differences in the 160/40 subset or by the choice of XGLM-564M as the multilingual evaluator."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Locale-conditioned rotating few-shot prompting eliminates demonstration regurgitation in 1.7B SLMs for PII substitution while producing more natural text than rule-based methods, though downstream NER training benefits more from synthetic variety than naturalness."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Locale-conditioned rotating few-shot prompts stop small language models from echoing demonstration examples during on-device PII substitution."}],"snapshot_sha256":"1767d0d1798e97fd46bb4e9176590148e9eb44ae983ad5d85fb458dc47ca6388"},"formal_canon":{"evidence_count":1,"snapshot_sha256":"13cea1224d6f07fff5395eb607961feb0d0eb9b58ef3be5215275e4827c88642"},"paper":{"abstract_excerpt":"Personally Identifiable Information (PII) redaction usually replaces detected entities with placeholder tokens such as [PERSON], destroying the downstream utility of the redacted text for retrieval and Named Entity Recognition (NER) training. We propose a fully on-device pipeline that substitutes PII with consistent, type-preserving fake values: a 1.5 B mixture-of-experts token classifier (openai/privacy-filter) detects spans, a 1-bit Bonsai-1.7B Small Language Model (SLM) proposes contextual surrogates for names, addresses, and dates, and a rule-based generator (faker) handles patterned field","authors_text":"Anuj Sadani, Deepak Kumar","cross_cats":["cs.AI"],"headline":"Locale-conditioned rotating few-shot prompts stop small language models from echoing demonstration examples during on-device PII substitution.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-13T13:47:11Z","title":"Locale-Conditioned Few-Shot Prompting Mitigates Demonstration Regurgitation in On-Device PII Substitution with Small Language Models"},"references":{"count":16,"internal_anchors":3,"resolved_work":16,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Faker: a Python package that generates fake data","work_id":"b1918f04-be18-4cdc-8b1b-8098eac3c9ec","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"llama.cpp: Port of LLaMA models in C/C++","work_id":"6b334f6b-fac2-4e0c-910e-43ed4197b5c2","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"spaCy: Industrial-strength natural language processing in Python","work_id":"8878703d-2330-47d2-81ec-6f3b4c7eacb6","year":2020},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Few-shot learning with multilingual language models","work_id":"32d56185-d394-41f6-a038-08c943a4650c","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Fantastically ordered prompts and where to find them: Overcoming few-shot prompt order sensitivity","work_id":"c6ecb82b-59ca-437f-91d8-470531f9f7c8","year":2022}],"snapshot_sha256":"8f7fc02398e1eadab2ea9d9fadc8e255a1580afbb705951483156df5e9d4b78b"},"source":{"id":"2605.13538","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T19:00:57.673618Z","id":"fddab2d7-c488-4d6a-8201-d6bc53eabbb7","model_set":{"reader":"grok-4.3"},"one_line_summary":"Locale-conditioned rotating few-shot prompting eliminates demonstration regurgitation in 1.7B SLMs for PII substitution while producing more natural text than rule-based methods, though downstream NER training benefits more from synthetic variety than naturalness.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Locale-conditioned rotating few-shot prompts stop small language models from echoing demonstration examples during on-device PII substitution.","strongest_claim":"With the fix, 482/482 unique Bonsai-1.7B calls succeed (no echoes) and produce locale-correct surrogates, although the SLM still copies from a small same-locale demonstration pool - a residual narrowness we quantify.","weakest_assumption":"That the observed NER performance gap is caused primarily by reduced variety in SLM outputs rather than by other unmeasured differences in the 160/40 subset or by the choice of XGLM-564M as the multilingual evaluator."}},"verdict_id":"fddab2d7-c488-4d6a-8201-d6bc53eabbb7"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:e205aef49527f7367491c5b765d6e765333d1476e8829e1eb8aab59b63cfaba6","target":"record","created_at":"2026-05-18T02:44:24Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"08982f296d3b7011c8b7b2fdf6bec4c5c580d305c1cc7386049148fae213f8dc","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-13T13:47:11Z","title_canon_sha256":"71ca1095213f9418fa258a120300c4beee4c761bb6f28c94731d6b266582527e"},"schema_version":"1.0","source":{"id":"2605.13538","kind":"arxiv","version":1}},"canonical_sha256":"d836d90d63182919648a0ead41da445340ce937cfd128a535c945fe435ff69ef","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d836d90d63182919648a0ead41da445340ce937cfd128a535c945fe435ff69ef","first_computed_at":"2026-05-18T02:44:24.088768Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:24.088768Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"UiiGXX66GqPyk1m3keJIB02P+V1lkUMgi/TVIISJ75wsBSzxgqypbEdhHkYG6O7UB9AclxzmKuuQDANrTUErDQ==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:24.089214Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13538","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e205aef49527f7367491c5b765d6e765333d1476e8829e1eb8aab59b63cfaba6","sha256:9aeb073cd6a4ee78316646a7a496d8c3908964698b74e9f10aa92be23816e870"],"state_sha256":"a43335df8b189df0ebc97c8c7478933f17a8b55e5f17e595cd88f3d5afcc5426"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1wAv2xrLfCLELwMKmKPUzbRfQ5n/wL98EgB0+rRoZz6iHpj+gxLPb2nGXfxKd5OjLHwepNCDjRzwYG0/sz5aDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-19T16:31:48.442104Z","bundle_sha256":"3343686f8dbd74e3954fb4312478e576288aaec2280c957e4db3864642c2a905"}}