{"paper":{"title":"crossfit: A Graph-Based Cross-Fitting Engine in R","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"The crossfit R package provides a general-purpose cross-fitting engine using user-specified DAGs of nuisance models with custom fold allocations.","cross_cats":[],"primary_cat":"stat.CO","authors_text":"Etienne Peyrot, Fran\\c{c}ois Petit","submitted_at":"2026-05-15T11:14:55Z","abstract_excerpt":"Cross-fitting is a key ingredient in many semiparametric estimation procedures, such as double/debiased machine learning (DML), enabling valid estimation of low-dimensional targets in the presence of high-dimensional nuisance functions by enforcing out-of-sample use of nuisance predictions. crossfit is an R package that provides a general-purpose, estimator-agnostic cross-fitting engine. Users specify (i) a target functional and (ii) a directed acyclic graph (DAG) of nuisance models, with node-specific training fold widths and target-specific evaluation windows. The engine executes a reproduci"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"crossfit is an R package that provides a general-purpose, estimator-agnostic cross-fitting engine. Users specify (i) a target functional and (ii) a directed acyclic graph (DAG) of nuisance models, with node-specific training fold widths and target-specific evaluation windows.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The package's internal scheduler and caching logic correctly execute the user-specified DAG and fold-allocation rules without introducing unintended data leakage or dependence between nuisance branches, as stated in the description of disjoint and independence-enforcing modes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"crossfit is an R package that supplies a general-purpose cross-fitting engine driven by user-specified DAGs of nuisance models with configurable fold allocations and reproducibility features.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"The crossfit R package provides a general-purpose cross-fitting engine using user-specified DAGs of nuisance models with custom fold allocations.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"c304fec5a643e53050b0bdf1ec2e782b72c891685f913ac54be714a7ce525bd1"},"source":{"id":"2605.15856","kind":"arxiv","version":1},"verdict":{"id":"fa9a1620-11ba-46b8-ad00-3dcee3c132ed","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T18:03:16.118805Z","strongest_claim":"crossfit is an R package that provides a general-purpose, estimator-agnostic cross-fitting engine. Users specify (i) a target functional and (ii) a directed acyclic graph (DAG) of nuisance models, with node-specific training fold widths and target-specific evaluation windows.","one_line_summary":"crossfit is an R package that supplies a general-purpose cross-fitting engine driven by user-specified DAGs of nuisance models with configurable fold allocations and reproducibility features.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The package's internal scheduler and caching logic correctly execute the user-specified DAG and fold-allocation rules without introducing unintended data leakage or dependence between nuisance branches, as stated in the description of disjoint and independence-enforcing modes.","pith_extraction_headline":"The crossfit R package provides a general-purpose cross-fitting engine using user-specified DAGs of nuisance models with custom fold allocations."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.15856/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T18:31:18.773741Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T18:10:15.757060Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T17:33:48.703845Z","status":"skipped","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T17:01:55.823110Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"53e50145ab7ff487023ce04a3d13a37cd2c094d7c660088fdb244f374ae541b9"},"references":{"count":15,"sample":[{"doi":"10.1214/aos/","year":1982,"title":"doi: 10.1214/aos/ 1176342360","work_id":"5bd55533-e64d-4884-b62b-14c429566078","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1214/aos/1176350055","year":1986,"title":"On Asymptotically Efficient Estimation in Semiparametric Models","work_id":"dec623fd-0fa8-4c67-bc86-16b9e1bc5fd2","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2011,"title":"Cross-Validated Targeted Minimum- Loss-Based Estimation","work_id":"90899aad-2f56-41ca-a394-455d3d3ebae3","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"arXiv preprint arXiv:1801.09138 , year=","work_id":"2a799d79-cace-45bb-98a8-de5ca7ea05f3","ref_index":4,"cited_arxiv_id":"1801.09138","is_internal_anchor":true},{"doi":"10.1111/ectj.12097","year":2018,"title":"Double/debiased machine learning for treatment and structural parameters.The Econometrics Journal, 21(1):C1–C68","work_id":"f1c0cebd-ff53-46ac-aa98-210211b3f171","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":15,"snapshot_sha256":"bbc0bf62988443cf2ed9f038af09d84a144abdcfea75d5ba5eecd318524a7163","internal_anchors":1},"formal_canon":{"evidence_count":1,"snapshot_sha256":"e73a6c647392ac05f4e8ba0795dd1dcc423d488ec9ac93b06ae50c549a15ac92"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}