{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:AMJ3ZQCEIHZ2JHWT2KL53BR7Z7","short_pith_number":"pith:AMJ3ZQCE","schema_version":"1.0","canonical_sha256":"0313bcc04441f3a49ed3d297dd863fcff4eb6b4c87d465a4428c522d74d8b83c","source":{"kind":"arxiv","id":"1906.01974","version":3},"attestation_state":"computed","paper":{"title":"Willump: A Statistically-Aware End-to-end Optimizer for Machine Learning Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DB","authors_text":"Daniel Kang, Deepak Narayanan, Matei Zaharia, Peter Bailis, Peter Kraft, Shoumik Palkar","submitted_at":"2019-06-03T22:43:00Z","abstract_excerpt":"Systems for ML inference are widely deployed today, but they typically optimize ML inference workloads using techniques designed for conventional data serving workloads and miss critical opportunities to leverage the statistical nature of ML. In this paper, we present Willump, an optimizer for ML inference that introduces two statistically-motivated optimizations targeting ML applications whose performance bottleneck is feature computation. First, Willump automatically cascades feature computation for classification queries: Willump classifies most data inputs using only high-value, low-cost f"},"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":"1906.01974","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DB","submitted_at":"2019-06-03T22:43:00Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3fe1bd822c3277e33617c781e099d3444c3f57080bde5b717741334b089043d4","abstract_canon_sha256":"cb7cbc1335c9e3c5c13d5154f34c2ec01b79f90158b5eaa513b9670044683799"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:45:51.197587Z","signature_b64":"ZxN+fC5BgDNofrY6mUMim57vr3WqZi9hnz2b15qggWOSkOQy9jNh2fEEzDSpuxxuXiYFnoDnK3UB7EA21L1nBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0313bcc04441f3a49ed3d297dd863fcff4eb6b4c87d465a4428c522d74d8b83c","last_reissued_at":"2026-07-05T00:45:51.197204Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:45:51.197204Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Willump: A Statistically-Aware End-to-end Optimizer for Machine Learning Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DB","authors_text":"Daniel Kang, Deepak Narayanan, Matei Zaharia, Peter Bailis, Peter Kraft, Shoumik Palkar","submitted_at":"2019-06-03T22:43:00Z","abstract_excerpt":"Systems for ML inference are widely deployed today, but they typically optimize ML inference workloads using techniques designed for conventional data serving workloads and miss critical opportunities to leverage the statistical nature of ML. In this paper, we present Willump, an optimizer for ML inference that introduces two statistically-motivated optimizations targeting ML applications whose performance bottleneck is feature computation. First, Willump automatically cascades feature computation for classification queries: Willump classifies most data inputs using only high-value, low-cost f"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1906.01974","kind":"arxiv","version":3},"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/1906.01974/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":"1906.01974","created_at":"2026-07-05T00:45:51.197262+00:00"},{"alias_kind":"arxiv_version","alias_value":"1906.01974v3","created_at":"2026-07-05T00:45:51.197262+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1906.01974","created_at":"2026-07-05T00:45:51.197262+00:00"},{"alias_kind":"pith_short_12","alias_value":"AMJ3ZQCEIHZ2","created_at":"2026-07-05T00:45:51.197262+00:00"},{"alias_kind":"pith_short_16","alias_value":"AMJ3ZQCEIHZ2JHWT","created_at":"2026-07-05T00:45:51.197262+00:00"},{"alias_kind":"pith_short_8","alias_value":"AMJ3ZQCE","created_at":"2026-07-05T00:45:51.197262+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.15381","citing_title":"DiffServe: Efficiently Serving Text-to-Image Diffusion Models with Query-Aware Model Scaling","ref_index":25,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7","json":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7.json","graph_json":"https://pith.science/api/pith-number/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/graph.json","events_json":"https://pith.science/api/pith-number/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/events.json","paper":"https://pith.science/paper/AMJ3ZQCE"},"agent_actions":{"view_html":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7","download_json":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7.json","view_paper":"https://pith.science/paper/AMJ3ZQCE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1906.01974&json=true","fetch_graph":"https://pith.science/api/pith-number/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/graph.json","fetch_events":"https://pith.science/api/pith-number/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/action/storage_attestation","attest_author":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/action/author_attestation","sign_citation":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/action/citation_signature","submit_replication":"https://pith.science/pith/AMJ3ZQCEIHZ2JHWT2KL53BR7Z7/action/replication_record"}},"created_at":"2026-07-05T00:45:51.197262+00:00","updated_at":"2026-07-05T00:45:51.197262+00:00"}