{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:27BXKOBSMWUUP6B4RVAF7YRQ3J","short_pith_number":"pith:27BXKOBS","canonical_record":{"source":{"id":"2404.04360","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-05T19:14:14Z","cross_cats_sorted":["cs.CL","cs.CR"],"title_canon_sha256":"b660bb358b105a49e42b78be2a43bf34c75daf4e57e9e221578444ebe686d755","abstract_canon_sha256":"48e838587d6970a8d49ff3f9763e85794b384295b30dc732df6c2707db7eaa8a"},"schema_version":"1.0"},"canonical_sha256":"d7c375383265a947f83c8d405fe230da6e34022e3eb38cfc624d05e70466e3c6","source":{"kind":"arxiv","id":"2404.04360","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.04360","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"arxiv_version","alias_value":"2404.04360v2","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.04360","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"pith_short_12","alias_value":"27BXKOBSMWUU","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"pith_short_16","alias_value":"27BXKOBSMWUUP6B4","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"pith_short_8","alias_value":"27BXKOBS","created_at":"2026-07-05T08:52:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:27BXKOBSMWUUP6B4RVAF7YRQ3J","target":"record","payload":{"canonical_record":{"source":{"id":"2404.04360","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-05T19:14:14Z","cross_cats_sorted":["cs.CL","cs.CR"],"title_canon_sha256":"b660bb358b105a49e42b78be2a43bf34c75daf4e57e9e221578444ebe686d755","abstract_canon_sha256":"48e838587d6970a8d49ff3f9763e85794b384295b30dc732df6c2707db7eaa8a"},"schema_version":"1.0"},"canonical_sha256":"d7c375383265a947f83c8d405fe230da6e34022e3eb38cfc624d05e70466e3c6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:52:53.125700Z","signature_b64":"8qJtUWXqFeBTp66P5FaH2mm8HqDIgYMJPjYMq5rW41iy29O+Cb1K1CmTzht70Tzu1Edbc0hMLmwYV82Wy7Z0DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d7c375383265a947f83c8d405fe230da6e34022e3eb38cfc624d05e70466e3c6","last_reissued_at":"2026-07-05T08:52:53.125232Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:52:53.125232Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2404.04360","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:52:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QgU1WIwJP/wxRbrpC4tL6miDhaSR1kN7DALPTg7rfzRWRW6TkwF+s5S73slosbFRoAAmJUwf1ql2YBGwU4OoCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T23:13:07.087266Z"},"content_sha256":"f1e9134f9ed6f2ff6894df46db399b2945097bfae2221e10ad4b44dd6572e055","schema_version":"1.0","event_id":"sha256:f1e9134f9ed6f2ff6894df46db399b2945097bfae2221e10ad4b44dd6572e055"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:27BXKOBSMWUUP6B4RVAF7YRQ3J","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Prompt Public Large Language Models to Synthesize Data for Private On-device Applications","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.CR"],"primary_cat":"cs.LG","authors_text":"Daniel Ramage, Shanshan Wu, Yanxiang Zhang, Yuanbo Zhang, Zheng Xu","submitted_at":"2024-04-05T19:14:14Z","abstract_excerpt":"Pre-training on public data is an effective method to improve the performance for federated learning (FL) with differential privacy (DP). This paper investigates how large language models (LLMs) trained on public data can improve the quality of pre-training data for the on-device language models trained with DP and FL. We carefully design LLM prompts to filter and transform existing public data, and generate new data to resemble the real user data distribution. The model pre-trained on our synthetic dataset achieves relative improvement of 19.0% and 22.8% in next word prediction accuracy compa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.04360","kind":"arxiv","version":2},"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/2404.04360/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:52:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sTNun9urO+LzCm0I2uyUbIApRi9xe25Bxi1fanHZba8smqiECQrh30tPVi6LSb60HSrEQLFb32G2O80GfEsQAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T23:13:07.088189Z"},"content_sha256":"dc5aa09dedf46deedcebeb32a555e524c8e180c43245033b4c96a4621d470db9","schema_version":"1.0","event_id":"sha256:dc5aa09dedf46deedcebeb32a555e524c8e180c43245033b4c96a4621d470db9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/27BXKOBSMWUUP6B4RVAF7YRQ3J/bundle.json","state_url":"https://pith.science/pith/27BXKOBSMWUUP6B4RVAF7YRQ3J/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/27BXKOBSMWUUP6B4RVAF7YRQ3J/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":"2026-08-08T23:13:07Z","links":{"resolver":"https://pith.science/pith/27BXKOBSMWUUP6B4RVAF7YRQ3J","bundle":"https://pith.science/pith/27BXKOBSMWUUP6B4RVAF7YRQ3J/bundle.json","state":"https://pith.science/pith/27BXKOBSMWUUP6B4RVAF7YRQ3J/state.json","well_known_bundle":"https://pith.science/.well-known/pith/27BXKOBSMWUUP6B4RVAF7YRQ3J/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:27BXKOBSMWUUP6B4RVAF7YRQ3J","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":"48e838587d6970a8d49ff3f9763e85794b384295b30dc732df6c2707db7eaa8a","cross_cats_sorted":["cs.CL","cs.CR"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-05T19:14:14Z","title_canon_sha256":"b660bb358b105a49e42b78be2a43bf34c75daf4e57e9e221578444ebe686d755"},"schema_version":"1.0","source":{"id":"2404.04360","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.04360","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"arxiv_version","alias_value":"2404.04360v2","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.04360","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"pith_short_12","alias_value":"27BXKOBSMWUU","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"pith_short_16","alias_value":"27BXKOBSMWUUP6B4","created_at":"2026-07-05T08:52:53Z"},{"alias_kind":"pith_short_8","alias_value":"27BXKOBS","created_at":"2026-07-05T08:52:53Z"}],"graph_snapshots":[{"event_id":"sha256:dc5aa09dedf46deedcebeb32a555e524c8e180c43245033b4c96a4621d470db9","target":"graph","created_at":"2026-07-05T08:52:53Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2404.04360/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Pre-training on public data is an effective method to improve the performance for federated learning (FL) with differential privacy (DP). This paper investigates how large language models (LLMs) trained on public data can improve the quality of pre-training data for the on-device language models trained with DP and FL. We carefully design LLM prompts to filter and transform existing public data, and generate new data to resemble the real user data distribution. The model pre-trained on our synthetic dataset achieves relative improvement of 19.0% and 22.8% in next word prediction accuracy compa","authors_text":"Daniel Ramage, Shanshan Wu, Yanxiang Zhang, Yuanbo Zhang, Zheng Xu","cross_cats":["cs.CL","cs.CR"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-05T19:14:14Z","title":"Prompt Public Large Language Models to Synthesize Data for Private On-device Applications"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.04360","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:f1e9134f9ed6f2ff6894df46db399b2945097bfae2221e10ad4b44dd6572e055","target":"record","created_at":"2026-07-05T08:52:53Z","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":"48e838587d6970a8d49ff3f9763e85794b384295b30dc732df6c2707db7eaa8a","cross_cats_sorted":["cs.CL","cs.CR"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-05T19:14:14Z","title_canon_sha256":"b660bb358b105a49e42b78be2a43bf34c75daf4e57e9e221578444ebe686d755"},"schema_version":"1.0","source":{"id":"2404.04360","kind":"arxiv","version":2}},"canonical_sha256":"d7c375383265a947f83c8d405fe230da6e34022e3eb38cfc624d05e70466e3c6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d7c375383265a947f83c8d405fe230da6e34022e3eb38cfc624d05e70466e3c6","first_computed_at":"2026-07-05T08:52:53.125232Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:52:53.125232Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8qJtUWXqFeBTp66P5FaH2mm8HqDIgYMJPjYMq5rW41iy29O+Cb1K1CmTzht70Tzu1Edbc0hMLmwYV82Wy7Z0DA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:52:53.125700Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.04360","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f1e9134f9ed6f2ff6894df46db399b2945097bfae2221e10ad4b44dd6572e055","sha256:dc5aa09dedf46deedcebeb32a555e524c8e180c43245033b4c96a4621d470db9"],"state_sha256":"d1ced3ddd5da27dae2d7175783cbd4e74b4bf01d8606f97a21582826f1719323"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2krDFkjl5umbAPB7oewkGufV02ksjQvrgl42XD4Mv7br29RsxDZ6CI/MPlddIOjU/H5NbB6NO8fx3Zyi4NuzAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T23:13:07.095172Z","bundle_sha256":"a7f716eaafa6f77d91608b19c82d916f8db2703911880418270ab001d3d9921a"}}