{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:ABUNJFO2O2HCWEP7BLUFAIX7QZ","short_pith_number":"pith:ABUNJFO2","schema_version":"1.0","canonical_sha256":"0068d495da768e2b11ff0ae85022ff867584da81d793ea693d7adbfce2f7b67e","source":{"kind":"arxiv","id":"2306.06076","version":2},"attestation_state":"computed","paper":{"title":"Differentially Private Image Classification by Learning Priors from Random Processes","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR","cs.LG","stat.ML"],"primary_cat":"cs.CV","authors_text":"Ashwinee Panda, Prateek Mittal, Vikash Sehwag, Xinyu Tang","submitted_at":"2023-06-08T04:14:32Z","abstract_excerpt":"In privacy-preserving machine learning, differentially private stochastic gradient descent (DP-SGD) performs worse than SGD due to per-sample gradient clipping and noise addition. A recent focus in private learning research is improving the performance of DP-SGD on private data by incorporating priors that are learned on real-world public data. In this work, we explore how we can improve the privacy-utility tradeoff of DP-SGD by learning priors from images generated by random processes and transferring these priors to private data. We propose DP-RandP, a three-phase approach. We attain new sta"},"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":"2306.06076","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-06-08T04:14:32Z","cross_cats_sorted":["cs.CR","cs.LG","stat.ML"],"title_canon_sha256":"ad1730d56444646d315f9d759ca93a14a2e4b17ad7a565950456e176ab005e92","abstract_canon_sha256":"2a5beb9ad15f0c1116c8730baa244c09d1853e2837b49870821fca5d25e407ef"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:07:11.987927Z","signature_b64":"yDpAum1CSAXCOgO7T3Ji4cBsntDnWa5anoVGMJyM/1yItKcm+vngcOA851LP9egXIBwpehPcAD/AgGSKs93kBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0068d495da768e2b11ff0ae85022ff867584da81d793ea693d7adbfce2f7b67e","last_reissued_at":"2026-07-05T07:07:11.986609Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:07:11.986609Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Differentially Private Image Classification by Learning Priors from Random Processes","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR","cs.LG","stat.ML"],"primary_cat":"cs.CV","authors_text":"Ashwinee Panda, Prateek Mittal, Vikash Sehwag, Xinyu Tang","submitted_at":"2023-06-08T04:14:32Z","abstract_excerpt":"In privacy-preserving machine learning, differentially private stochastic gradient descent (DP-SGD) performs worse than SGD due to per-sample gradient clipping and noise addition. A recent focus in private learning research is improving the performance of DP-SGD on private data by incorporating priors that are learned on real-world public data. In this work, we explore how we can improve the privacy-utility tradeoff of DP-SGD by learning priors from images generated by random processes and transferring these priors to private data. We propose DP-RandP, a three-phase approach. We attain new sta"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.06076","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/2306.06076/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":"2306.06076","created_at":"2026-07-05T07:07:11.987385+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.06076v2","created_at":"2026-07-05T07:07:11.987385+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.06076","created_at":"2026-07-05T07:07:11.987385+00:00"},{"alias_kind":"pith_short_12","alias_value":"ABUNJFO2O2HC","created_at":"2026-07-05T07:07:11.987385+00:00"},{"alias_kind":"pith_short_16","alias_value":"ABUNJFO2O2HCWEP7","created_at":"2026-07-05T07:07:11.987385+00:00"},{"alias_kind":"pith_short_8","alias_value":"ABUNJFO2","created_at":"2026-07-05T07:07:11.987385+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/ABUNJFO2O2HCWEP7BLUFAIX7QZ","json":"https://pith.science/pith/ABUNJFO2O2HCWEP7BLUFAIX7QZ.json","graph_json":"https://pith.science/api/pith-number/ABUNJFO2O2HCWEP7BLUFAIX7QZ/graph.json","events_json":"https://pith.science/api/pith-number/ABUNJFO2O2HCWEP7BLUFAIX7QZ/events.json","paper":"https://pith.science/paper/ABUNJFO2"},"agent_actions":{"view_html":"https://pith.science/pith/ABUNJFO2O2HCWEP7BLUFAIX7QZ","download_json":"https://pith.science/pith/ABUNJFO2O2HCWEP7BLUFAIX7QZ.json","view_paper":"https://pith.science/paper/ABUNJFO2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.06076&json=true","fetch_graph":"https://pith.science/api/pith-number/ABUNJFO2O2HCWEP7BLUFAIX7QZ/graph.json","fetch_events":"https://pith.science/api/pith-number/ABUNJFO2O2HCWEP7BLUFAIX7QZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ABUNJFO2O2HCWEP7BLUFAIX7QZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ABUNJFO2O2HCWEP7BLUFAIX7QZ/action/storage_attestation","attest_author":"https://pith.science/pith/ABUNJFO2O2HCWEP7BLUFAIX7QZ/action/author_attestation","sign_citation":"https://pith.science/pith/ABUNJFO2O2HCWEP7BLUFAIX7QZ/action/citation_signature","submit_replication":"https://pith.science/pith/ABUNJFO2O2HCWEP7BLUFAIX7QZ/action/replication_record"}},"created_at":"2026-07-05T07:07:11.987385+00:00","updated_at":"2026-07-05T07:07:11.987385+00:00"}