{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:W2IXFVWCDC76WC66NGC7EN5L3J","short_pith_number":"pith:W2IXFVWC","schema_version":"1.0","canonical_sha256":"b69172d6c218bfeb0bde6985f237abda4a5dfe584b8ae68d06f503a6586f5fc5","source":{"kind":"arxiv","id":"2605.14578","version":1},"attestation_state":"computed","paper":{"title":"Woodelf++: A Fast and Unified Partial Dependence Plot Algorithm for Decision Tree Ensembles","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Woodelf++ computes partial dependence plots, joint plots, and all-order interaction values for decision tree ensembles with exponential complexity reductions.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexander Nadel, Ron Wettenstein, Udi Boker","submitted_at":"2026-05-14T08:49:19Z","abstract_excerpt":"Partial Dependence Plots (PDPs) visualize how changes in a single feature affect the average model prediction. They are widely used in practice to interpret decision tree ensembles and other machine learning models. Joint-PDPs extend this idea to pairs of features, revealing their combined effect. Partial Dependence Interaction Values (PDIVs) measure feature interactions. The Any-Order-PDIVs task computes these interactions for every feature subset across all rows of the dataset.\n  We introduce Woodelf++, a unified and efficient approach for computing all these useful explainability tools on d"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.14578","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-14T08:49:19Z","cross_cats_sorted":[],"title_canon_sha256":"9e01914691aaefb046cf4766510c703b04189bc2a9d28c747de421bd51546420","abstract_canon_sha256":"fcffcd67b46807ee09eba5fd1ce5fd7aae4b07e053a9b362afd7ae821841f87c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:05.396843Z","signature_b64":"hjUjnL9A1AnOhm/NabHd+IhHpx0YSOpf+qiFhjeTo0tnLToAtzL237BpHstgdDOuarVwYs8IJzPcESuaZ2ESDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b69172d6c218bfeb0bde6985f237abda4a5dfe584b8ae68d06f503a6586f5fc5","last_reissued_at":"2026-05-17T23:39:05.396192Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:05.396192Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Woodelf++: A Fast and Unified Partial Dependence Plot Algorithm for Decision Tree Ensembles","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Woodelf++ computes partial dependence plots, joint plots, and all-order interaction values for decision tree ensembles with exponential complexity reductions.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexander Nadel, Ron Wettenstein, Udi Boker","submitted_at":"2026-05-14T08:49:19Z","abstract_excerpt":"Partial Dependence Plots (PDPs) visualize how changes in a single feature affect the average model prediction. They are widely used in practice to interpret decision tree ensembles and other machine learning models. Joint-PDPs extend this idea to pairs of features, revealing their combined effect. Partial Dependence Interaction Values (PDIVs) measure feature interactions. The Any-Order-PDIVs task computes these interactions for every feature subset across all rows of the dataset.\n  We introduce Woodelf++, a unified and efficient approach for computing all these useful explainability tools on d"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Woodelf++ computes Any-Order-PDIVs for every feature subset across all rows with an exponential complexity improvement, finishing in 5 minutes on a 400k-row dataset while the state of the art would require over 1,000,000 years.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The claimed exponential gain and practical speedups rest on the assumption that the pseudo-Boolean function metrics derived from the tree ensemble can be evaluated without hidden per-row or per-subset overheads that grow with dataset size or number of features.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Woodelf++ computes exact and approximate PDPs, Joint-PDPs, and Any-Order-PDIVs for decision tree ensembles with up to exponential speedups by deriving metrics over pseudo-Boolean functions.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Woodelf++ computes partial dependence plots, joint plots, and all-order interaction values for decision tree ensembles with exponential complexity reductions.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"d1fb408652e199b7eab342b26948ef25142e30dd0a9c9c05a8e733b44adc770e"},"source":{"id":"2605.14578","kind":"arxiv","version":1},"verdict":{"id":"2d80a408-80ef-49bb-89f3-25974a30a2c0","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T01:37:51.929724Z","strongest_claim":"Woodelf++ computes Any-Order-PDIVs for every feature subset across all rows with an exponential complexity improvement, finishing in 5 minutes on a 400k-row dataset while the state of the art would require over 1,000,000 years.","one_line_summary":"Woodelf++ computes exact and approximate PDPs, Joint-PDPs, and Any-Order-PDIVs for decision tree ensembles with up to exponential speedups by deriving metrics over pseudo-Boolean functions.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The claimed exponential gain and practical speedups rest on the assumption that the pseudo-Boolean function metrics derived from the tree ensemble can be evaluated without hidden per-row or per-subset overheads that grow with dataset size or number of features.","pith_extraction_headline":"Woodelf++ computes partial dependence plots, joint plots, and all-order interaction values for decision tree ensembles with exponential complexity reductions."},"references":{"count":27,"sample":[{"doi":"","year":2026,"title":"Why Should I Trust You?","work_id":"d5d8ac43-fb9f-44d2-a4bd-4894f421f1f0","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Setp[i] = 1ifx i ∈S + k","work_id":"56ec3976-c5f3-4bcf-8ee1-a6bc3f36a7d1","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Setp[i] = 2if(x i ∈S − k )∧(x i ∈s)","work_id":"924dae2f-8131-4312-86f7-03728bd3f1d8","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Setp[i] = 3if(x i ∈S − k )∧(x i /∈s)","work_id":"bfe910a4-3bb6-401d-a8fc-dd3bef9e399b","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"In par- ticular, there is noisuch thatx i ∈S + k andx i ∈S − k","work_id":"0e3c33e0-ce88-496e-a3de-5ee264e29efd","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":27,"snapshot_sha256":"c2349fc88b052450c25f44611eaf93b37c590473f39ab73b8270d7b9bca2f0a2","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ee4ae89f860ebfb44380cec0273cdd9dc9cd421c0faf0c9252398734de9acd06"},"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":"2605.14578","created_at":"2026-05-17T23:39:05.396288+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.14578v1","created_at":"2026-05-17T23:39:05.396288+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14578","created_at":"2026-05-17T23:39:05.396288+00:00"},{"alias_kind":"pith_short_12","alias_value":"W2IXFVWCDC76","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"W2IXFVWCDC76WC66","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"W2IXFVWC","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J","json":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J.json","graph_json":"https://pith.science/api/pith-number/W2IXFVWCDC76WC66NGC7EN5L3J/graph.json","events_json":"https://pith.science/api/pith-number/W2IXFVWCDC76WC66NGC7EN5L3J/events.json","paper":"https://pith.science/paper/W2IXFVWC"},"agent_actions":{"view_html":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J","download_json":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J.json","view_paper":"https://pith.science/paper/W2IXFVWC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.14578&json=true","fetch_graph":"https://pith.science/api/pith-number/W2IXFVWCDC76WC66NGC7EN5L3J/graph.json","fetch_events":"https://pith.science/api/pith-number/W2IXFVWCDC76WC66NGC7EN5L3J/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J/action/timestamp_anchor","attest_storage":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J/action/storage_attestation","attest_author":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J/action/author_attestation","sign_citation":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J/action/citation_signature","submit_replication":"https://pith.science/pith/W2IXFVWCDC76WC66NGC7EN5L3J/action/replication_record"}},"created_at":"2026-05-17T23:39:05.396288+00:00","updated_at":"2026-05-17T23:39:05.396288+00:00"}